Algebraic Topology Homework 2

import%20marimo%0A%0A__generated_with%20%3D%20%220.17.6%22%0Aapp%20%3D%20marimo.App()%0A%0A%0A%40app.cell%0Adef%20_()%3A%0A%20%20%20%20import%20marimo%20as%20mo%0A%20%20%20%20return%20(mo%2C)%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20nav_menu%20%3D%20mo.nav_menu(%0A%20%20%20%20%20%20%20%20%7B%0A%20%20%20%20%20%20%20%20%20%20%20%20%22%2Findex.html%22%3A%20f%22%7Bmo.icon('lucide%3Ahome')%7D%20Home%22%2C%0A%20%20%20%20%20%20%20%20%7D%2C%0A%20%20%20%20%20%20%20%20orientation%3D%22vertical%22%2C%0A%20%20%20%20)%0A%20%20%20%20nav_menu%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%20Spring%20Semester%20Homework%201%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%23%20Problem%2018%0A%0A%20%20%20%20**For%20the%20closed%20orientable%20surface%20%24M%24%20of%20genus%20%24g%5Cge%201%24%2C%20show%20that%20for%20each%20non-zero%20%24%5Calpha%20%5Cin%20H%5E1(M%3B%5Cmathbb%7BZ%7D)%24%20there%20exists%20%24%5Cbeta%20%5Cin%20H%5E1(M%3B%5Cmathbb%7BZ%7D)%24%20with%20%24%5Calpha%20%5Cbeta%20%5Cne%200%24.%20Deduce%20that%20%24M%24%20is%20not%20homotopy%20equivalent%20to%20a%20wedge%20sum%20%24X%20%5Cvee%20Y%24%20of%20CW-complexes%20with%20non-trivial%20reduced%20homology.%20Do%20the%20same%20for%20closed%20non-orientable%20surfaces%20using%20cohomology%20with%20%24%5Cmathbb%7BZ%7D_2%24%20coefficients**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230121190401.png%22%20width%3D%22300%22%20%2F%3E%0A%0A%20%20%20%20Examine%20Example%203.7.%20With%20the%20usual%20%24%5CDelta%24-complex%20structure%20we%20have%20for%20%24M%24%2C%20we%20have%20that%20the%20cup%20product%20of%20(cohomology%20classes)%20%24%5Calpha_i%24%20and%20%24%5Cbeta_j%24%20is%20%24%5Calpha_i%20%5Ccup%20%5Cbeta_j%20%3D%20%5Cbegin%7Bcases%7D%5Cgamma%2C%26i%20%3D%20j%5C%5C%200%2C%26i%5Cne%20j%5Cend%7Bcases%7D%24%20for%20%24%5Cgamma%24%20being%20the%20dual%20generator%20of%20%24H%5E2(M%3B%5Cmathbb%7BZ%7D)%24.%20In%20particular%2C%20for%20any%20%24%5Calpha%24%20that%20equals%20%24%5Calpha_i%24%20or%20%24%5Cbeta_j%24%2C%20we%20can%20easily%20find%20%24%5Cbeta%24%20such%20that%20%24%5Calpha%20%5Cbeta%20%5Cne%200%24.%20Since%20we%20know%20that%20%24%5Calpha_i%24%20and%20%24%5Cbeta_j%24%20form%20a%20basis%20for%20%24H%5E1(M%3B%20%5Cmathbb%7BZ%7D)%24%20and%20cup%20product%20is%20distributive%2C%20the%20desired%20statement%20also%20works%20for%20any%20%24%5Calpha%20%5Cin%20H%5E1(M%3B%20%5Cmathbb%7BZ%7D)%24.%0A%0A%20%20%20%20Suppose%20%24M%24%20is%20homotopy%20equivalent%20to%20a%20wedge%20sum%20%24X%20%5Cvee%20Y%24%2C%20then%20their%20reduced%20(co)homology%20groups%20are%20isomorphic.%20In%20particular%2C%20%24H%5E2(X%20%5Cvee%20Y%3B%20%5Cmathbb%7BZ%7D)%20%5Ccong%24%20%24H%5E2(M%3B%20%5Cmathbb%7BZ%7D)%20%5Ccong%24%20%24%5Cmathbb%7BZ%7D%24.%20Since%20%24H%5E2(X%20%5Cvee%20Y%3B%20%5Cmathbb%7BZ%7D)%20%3D%20H%5E2(X%3B%20%5Cmathbb%7BZ%7D)%20%5Ctimes%20H%5E2(Y%3B%5Cmathbb%7BZ%7D)%24%2C%20we%20must%20have%20one%20of%20%24H%5E2(X%3B%20%5Cmathbb%7BZ%7D)%24%20and%20%24H%5E2(Y%3B%5Cmathbb%7BZ%7D)%24%20is%20%24%5Cmathbb%7BZ%7D%24%20and%20the%20other%20is%20%240%24.%0A%0A%20%20%20%20WLOG%2C%20say%20%24H%5E2(X%3B%20%5Cmathbb%7BZ%7D)%20%3D%200%24.%20With%20similar%20argument%2C%20it%20is%20easy%20to%20see%20%24H%5En(X%3B%20%5Cmathbb%7BZ%7D)%20%3D%200%24%20for%20all%20%24n%20%3E%202%24%20and%20%24n%20%3D%200%24.%20But%20we%20assume%20%24X%24%20has%20non-trivial%20reduced%20homology%2C%20thus%20%24H%5E1(X%3B%20%5Cmathbb%7BZ%7D)%24%20must%20not%20be%20%240%24.%20Thus%20we%20can%20pick%20some%20%24%5Calpha%20%5Cin%20H%5E1(X%3B%20%5Cmathbb%7BZ%7D)%24%20that%20is%20non-zero.%0A%0A%20%20%20%20Let%20%24%5Cbeta%20%5Cin%20H%5E1(X%20%5Cvee%20Y%3B%20%5Cmathbb%7BZ%7D)%24%20be%20any%20element%20and%20consider%20inclusions%20%24i_X%3A%20X%20%5Cto%20X%20%5Cvee%20Y%24%2C%20%24i_Y%3A%20Y%20%5Cto%20X%20%5Cvee%20Y%24%20which%20induce%20%24i_X%5E*%3A%20H%5E*(X%20%5Cvee%20Y%3B%20%5Cmathbb%7BZ%7D)%20%5Cto%20H%5E*(X%3B%20%5Cmathbb%7BZ%7D)%24%20and%20%24i_Y%5E*%3A%20H%5E*(X%20%5Cvee%20Y%3B%20%5Cmathbb%7BZ%7D)%20%5Cto%20H%5E*(Y%3B%20%5Cmathbb%7BZ%7D)%24%3A%0A%0A%20%20%20%201.%20Since%20%24H%5E2(X%3B%20%5Cmathbb%7BZ%7D)%20%3D%200%24%2C%20we%20always%20have%20%24H%5E2(X%3B%20%5Cmathbb%7BZ%7D)%20%5Cni%20i_X%5E*(%5Calpha%5Cbeta)%20%3D%200%24%3B%0A%20%20%20%202.%20Since%20%24%5Calpha%20%5Cin%20H%5E1(X%3B%20%5Cmathbb%7BZ%7D)%24%2C%20we%20have%20%24i_Y%5E*(%5Calpha)%3D0%24%20thus%20%24i_Y%5E*(%5Calpha%5Cbeta)%20%3D%200%24%20because%20it's%20a%20ring%20homomorphism.%0A%0A%20%20%20%20Since%20%24i_X%5E*%24%20and%20%24i_Y%5E*%24%20induce%20isomorphism%20%24H%5E2(X%20%5Cvee%20Y%3B%20%5Cmathbb%7BZ%7D)%20%5Ccong%20H%5E2(X%3B%20%5Cmathbb%7BZ%7D)%20%5Ctimes%20H%5E2(Y%3B%20%5Cmathbb%7BZ%7D)%24%2C%20and%20the%20image%20of%20%24%5Calpha%5Cbeta%24%20under%20this%20isomorphism%20is%20%240%24%2C%20we%20must%20have%20that%20%24%5Calpha%5Cbeta%20%5Cin%20H%5E2(X%20%5Cvee%20Y%3B%20%5Cmathbb%7BZ%7D)%24%20is%20%240%24.%0A%0A%20%20%20%20But%20we%20also%20have%20that%20%24H%5E2(X%20%5Cvee%20Y%3B%20%5Cmathbb%7BZ%7D)%24%20is%20isomorphic%20to%20%24H%5E2(M%3B%20%5Cmathbb%7BZ%7D)%24%2C%20thus%20we%20have%20find%20some%20non-zero%20%24%5Calpha_M%20%5Cin%20H%5E1(M%3B%20%5Cmathbb%7BZ%7D)%24%20such%20that%20%24%5Calpha_M%20%5Cbeta_M%20%3D%200%24%20for%20all%20%24%5Cbeta_M%20%5Cin%20H%5E1(M%3B%20%5Cmathbb%7BZ%7D)%24%2C%20which%20contradict%20the%20statement%20we%20have%20above.%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230122120532.png%22%20width%3D%22300%22%20%2F%3E%0A%0A%20%20%20%20For%20the%20cohomology%20of%20non-orientable%20surface%20with%20%24%5Cmathbb%7BZ%7D_2%24%20coefficients%2C%20examine%20Example%203.8%20with%20the%20%24%5CDelta%24-complex%20structure%20on%20the%20right%2C%20then%20we%20know%20the%20product%20%24%5Calpha_i%20%5Calpha_i%24%20is%20non-zero.%20Since%20%24%5Calpha_i%24's%20form%20a%20basis%20for%20%24H%5E1(M%3B%20%5Cmathbb%7BZ%7D_2)%24%2C%20we%20have%20the%20desired%20statement.%0A%0A%20%20%20%20For%20the%20second%20part%2C%20since%20for%20a%20non-orientable%20surface%20(so%20dimension%20is%20%242%24)%2C%20we%20must%20have%20that%20%24H%5E2(M%3B%20%5Cmathbb%7BZ%7D_2)%20%5Ccong%20%5Cmathbb%7BZ%7D_2%24.%20So%20that%20if%20%24M%24%20is%20homotopy%20equivalent%20with%20%24X%20%5Cvee%20Y%24%20then%20we%20must%20have%20that%20exactly%20one%20of%20%24H%5E2(X%3B%20%5Cmathbb%7BZ%7D_2)%24%20and%20%24H%5E2(Y%3B%20%5Cmathbb%7BZ%7D_2)%24%20is%20%24%5Cmathbb%7BZ%7D_2%24%2C%20and%20we%20can%20follow%20the%20same%20argument%20to%20get%20the%20result.%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%20Spring%20Semester%20Homework%202%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%23%20Problem%2011%0A%0A%20%20%20%20**If%20%24M_g%24%20denotes%20the%20closed%20orientable%20surface%20of%20genus%20%24g%24%2C%20show%20that%20degree%20%241%24%20maps%20%24M_g%20%5Cto%20M_h%24%20exists%20iff%20%24g%20%5Cge%20h%24**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Suppose%20%24g%5Cge%20h%24%2C%20consider%20the%20map%20that%20maps%20%24M_%7Bg'%7D%24%2C%20the%20interior%20of%20a%20surface%20of%20genus%20%24h%24%20with%20one%20boundary%20circle%2C%20into%20%24M_h%24%20minus%20a%20point%20%24x%24%20in%20a%20natural%20way%2C%20and%20that%20maps%20the%20rest%20(%24M_g%20%5Csetminus%20M_%7Bg'%7D%24)%20to%20the%20single%20point%20%24x%24.%20This%20map%20has%20%24x%24%20as%20the%20only%20critical%20point%20and%20has%20degree%20%241%24.%0A%0A%20%20%20%20Suppose%20%24g%20%3C%20h%24.%20Suppose%20a%20degree%20%241%24%20map%20%24f%3A%20M_g%20%5Cto%20M_h%24%20exists%2C%20then%20by%20the%20previous%20question%2C%20%24f_*%3A%20H_1(M_g)%20%5Ccong%20%5Cmathbb%7BZ%7D%5E%7B2g%7D%20%5Cto%20H_1(M_h)%20%5Ccong%20%5Cmathbb%7BZ%7D%5E%7B2h%7D%24%20is%20surjective%2C%20which%20is%20impossible.%20So%20the%20other%20direction%20also%20holds%20and%20we%20get%20the%20if%20and%20only%20if%20statement.%0A%0A%20%20%20%20%23%23%20Problem%2012%0A%0A%20%20%20%20**As%20an%20algebraic%20application%20of%20the%20preceding%20problem%2C%20show%20that%20in%20a%20free%20group%20%24F%24%20with%20basis%20%24x_1%2C%20%5Cdots%2C%20x_%7B2k%7D%24%2C%20the%20product%20of%20commutators%20%24%5Bx_1%2Cx_2%5D%5Cdots%5Bx_%7B2k-1%7D%2C%20x_%7B2k%7D%5D%24%20is%20not%20equal%20to%20a%20product%20of%20fewer%20than%20%24k%24%20commutators%20%24%5Bv_i%2C%20w_i%5D%24%20of%20elements%20%24v_i%2C%20w_i%20%5Cin%20F%24**.%20*Hint%3A%20Recall%20that%20the%20%242%24-ell%20of%20%24M_k%24%20is%20attached%20by%20the%20product%20%24%5Bx_1%2Cx_2%5D%5Cdots%5Bx_%7B2k-1%7D%2Cx_%7B2k%7D%5D%24.%20From%20a%20relation%20%24%5Bx_1%2Cx_2%5D%5Cdots%5Bx_%7B2k-1%7D%2Cx_%7B2k%7D%5D%24%20%24%3D%20%5Bv_1%2Cw_1%5D%5Cdots%5Bv_j%2Cw_j%5D%24%20in%20%24F%24%2C%20construct%20a%20degree%20%241%24%20map%20%24M_j%20%5Cto%20M_k%24*.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Consider%20CW-complex%20structures%20on%20%24M_j%24%20(with%20edges%20%24a_1%2C%5Cdots%2Ca_%7B2j%7D%24)%20and%20%24M_k%24%20(with%20edges%20%24x_1%2C%5Cdots%2Cx_%7B2k%7D%24)%20of%20genus%20%24j%2C%20k%24%20closed%20orientable%20surfaces.%20Define%20a%20map%20%24f%24%20maps%20between%20the%20%241%24-skeletons%2C%20such%20that%20%24a_%7B2i-1%7D%24%20is%20mapped%20to%20%24v_i%24%2C%20and%20%24a_%7B2i%7D%24%20is%20mapped%20to%20%24w_i%24.%20Then%20the%20product%20%24%5Ba_1%2Ca_2%5D%5Cdots%20%5Ba_%7B2j-1%7D%2Ca_%7B2j%7D%5D%24%20is%20mapped%20to%20%24%5Bv_1%2Cw_1%5D%5Cdots%5Bv_j%2Cw_j%5D%24%20which%20by%20assumption%20equals%20%24%5Bx_1%2Cx_2%5D%5Cdots%5Bx_%7B2k-1%7D%2Cx_%7B2k%7D%5D%24.%0A%0A%20%20%20%20Extend%20%24f%24%20to%20%24F%3A%20M_j%20%5Cto%20M_k%24%2C%20since%20the%20edges%20are%20mapped%20as%20above%2C%20the%20%242%24-cell%20of%20%24M_j%24%20is%20mapped%20(homeomorphically)%20to%20the%20%242%24-cell%20of%20%24M_k%24%2C%20thus%20%24F%24%20has%20degree%20%241%24%2C%20but%20then%20the%20previous%20question%20gives%20a%20contradiction%20because%20%24j%20%5Cnot%5Cge%20k%24.%0A%0A%20%20%20%20%23%23%20Problem%2013%0A%0A%20%20%20%20**Let%20%24M_h'%20%5Csubset%20M_g%24%20be%20a%20compact%20subsurface%20of%20genus%20%24h%24%20with%20one%20boundary%20circle%2C%20so%20%24M_h'%24%20is%20homeomorphic%20to%20%24M_h%24%20with%20an%20open%20disk%20removed.%20Show%20there%20is%20no%20retraction%20%24M_g%20%5Cto%20M_h'%24%20if%20%24h%20%3E%20g%2F2%24**.%20*Hint%3A%20Apply%20the%20previous%20problem%2C%20using%20the%20fact%20that%20%24M_g%20-%20M_h'%24%20has%20genus%20%24g%20-%20h%24*.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Assume%20we%20have%20a%20retraction%20%24r%3A%20M_g%20%5Cto%20M_h'%24.%20Consider%20the%20surface%20%24%5Coverline%7BM_g%20-%20M_h'%7D%24%2C%20it%20has%20genus%20%24g%20-%20h%24%20with%20one%20boundary%20circle.%20So%20we%20can%20define%20a%20map%20%24i%3A%20M_%7Bg-h%7D%20%5Cto%20%5Coverline%7BM_g%20-%20M_h'%7D%24%20where%20it%20maps%20all%20point%20but%20one%20naturally%20to%20%24%5Coverline%7BM_g%20-%20M_h'%7D%24%2C%20and%20maps%20the%20remaining%20one%20point%20to%20an%20arbitrary%20point%20on%20the%20boundary%20circle%20of%20%24%5Coverline%7BM_g%20-%20M_h'%7D%24.%0A%0A%20%20%20%20Consider%20%24r%20%5Ccirc%20i%3A%20M_%7Bg-h%7D%20%5Cto%20%5Coverline%7BM_g%20-%20M_h'%7D%20%5Cto%20M_h'%24.%20We%20have%20%24%5Cdeg(r)%20%3D%201%24%20because%20it%20is%20a%20retraction%2C%20and%20%24%5Cdeg(i)%20%3D%201%24%20because%20it%20can%20be%20understood%20as%20an%20inclusion.%20Thus%20%24%5Cdeg(r%20%5Ccirc%20i)%20%3D%201%24%20but%20%24g%20-%20h%3C%20h%24%20because%20%24g%3C2h%24%2C%20thus%20we%20have%20a%20contradiction%20from%20question%2011.%0A%0A%20%20%20%20%23%23%20Problem%2024%0A%0A%20%20%20%20**Let%20%24M%24%20be%20a%20closed%20connected%20%243%24-manifold%2C%20and%20write%20%24H_1(M%3B%5Cmathbb%7BZ%7D)%24%20as%20%24%5Cmathbb%7BZ%7D%5Er%20%5Coplus%20F%24%2C%20the%20direct%20sum%20of%20a%20free%20abelian%20group%20of%20rank%20%24r%24%20and%20a%20finite%20group%20%24F%24.%20Show%20that%20%24H_2(M%3B%20%5Cmathbb%7BZ%7D)%24%20is%20%24%5Cmathbb%7BZ%7D%5Er%24%20if%20%24M%24%20is%20orientable%20and%20%24%5Cmathbb%7BZ%7D%5E%7Br-1%7D%20%5Coplus%20%5Cmathbb%7BZ%7D_2%24%20if%20%24M%24%20is%20non-orientable**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%201.%20Suppose%20%24M%24%20is%20orientable.%20We%20know%20that%20%24H_3(M%3B%5Cmathbb%7BZ%7D)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%20(orientable%2C%20dimension%20%243%24)%2C%20%24H_0(M%3B%20%5Cmathbb%7BZ%7D)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%20(connected)%2C%20and%20%24H_i(M%3B%5Cmathbb%7BZ%7D)%20%3D%200%24%20whenever%20%24i%3E3%24%20(dimension%20%243%24).%20By%20Corollary%203.37%2C%20we%20have%20that%20%24%5Cchi(M)%20%3D%200%24.%20Recall%20now%20that%20the%20Euler%20characteristic%20is%20the%20alternating%20sum%20of%20rank%20of%20homology%20groups%3A%20since%20we%20assume%20%24H_1(M%3B%5Cmathbb%7BZ%7D)%24%20has%20rank%20%24r%24%2C%20we%20must%20have%20%24H_2(M%3B%20%5Cmathbb%7BZ%7D)%24%20also%20have%20rank%20%24r%24.%20By%20Corollary%203.28%2C%20%24H_%7B3-1%7D(M%3B%20%5Cmathbb%7BZ%7D)%24%20is%20torsion-free%2C%20thus%20%24H_2(M%3B%20%5Cmathbb%7BZ%7D)%20%3D%20%5Cmathbb%7BZ%7D%5Er%24%3B%0A%20%20%20%202.%20Suppose%20%24M%24%20is%20non-orientable.%20Follow%20the%20same%20argument%2C%20this%20time%20we%20have%20%24H_3(M%3B%20%5Cmathbb%7BZ%7D)%20%5Ccong%200%24%2C%20thus%20%24H_2(M%3B%20%5Cmathbb%7BZ%7D)%24%20has%20rank%20%24r-1%24%20by%20Corollary%203.37.%20Again%20by%20Corollary%203.28%2C%20we%20get%20that%20torsion%20of%20%24H_2(M%3B%20%5Cmathbb%7BZ%7D)%24%20is%20%24%5Cmathbb%7BZ%7D_2%24%20thus%20we%20have%20%24H_2(M%3B%5Cmathbb%7BZ%7D)%20%5Ccong%20%5Cmathbb%7BZ%7D%5E%7Br-1%7D%20%5Coplus%20%5Cmathbb%7BZ%7D_2%24.%0A%0A%20%20%20%20%23%23%20Problem%2025%0A%0A%20%20%20%20**Show%20that%20if%20a%20closed%20orientable%20manifold%20%24M%24%20of%20dimension%20%242k%24%20has%20%24H_%7Bk-1%7D(M%3B%5Cmathbb%7BZ%7D)%24%20torsion-free%2C%20then%20%24H_k(M%3B%20%5Cmathbb%7BZ%7D)%24%20is%20also%20torsion-free**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20The%20UCT%20for%20cohomology%20says%20that%20the%20following%20exact%20sequence%20splits%3A%20%24%240%20%5Cto%20%5Ctext%7BExt%7D(H_%7Bn-1%7D(M%3B%5Cmathbb%7BZ%7D)%2C%5Cmathbb%7BZ%7D)%20%5Cto%20H%5En(M%3B%5Cmathbb%7BZ%7D)%20%5Cto%20%5Ctext%7BHom%7D(H_n(M%3B%5Cmathbb%7BZ%7D)%2C%5Cmathbb%7BZ%7D)%20%5Cto%200%2C%24%24%20where%20%24%5Ctext%7BExt%7D(H%2C%20%5Cmathbb%7BZ%7D)%24%20is%20isomorphic%20to%20the%20torsion%20subgroup%20of%20%24H%24%20(if%20it%20is%20finitely%20generated)%2C%20and%20%24%5Ctext%7BHom%7D(H%2C%20%5Cmathbb%7BZ%7D)%24%20is%20isomorphic%20to%20certain%20%24%5Cmathbb%7BZ%7D%5Eb%24%20where%20%24b%24%20is%20the%20Betti%20number.%20Thus%20we%20have%20that%20%24H%5Ek(M%3B%5Cmathbb%7BZ%7D)%20%5Ccong%20%5Cmathbb%7BZ%7D%5Eb%20%5Coplus%20T_%7Bk-1%7D%24%20(take%20%24n%20%3D%20k%24).%0A%0A%20%20%20%20Poincar%C3%A9%20Duality%20says%20that%20%24H%5Ek(M%3B%5Cmathbb%7BZ%7D)%20%5Ccong%20H_%7B2k-k%20%3D%20k%7D(M%3B%20%5Cmathbb%7BZ%7D)%24.%20Now%20if%20%24T_%7Bk-1%7D%24%20is%20trivial%20as%20assumed%2C%20then%20%24H%5Ek(M%3B%5Cmathbb%7BZ%7D)%20%5Ccong%20%5Cmathbb%7BZ%7D%5Eb%24%20%24%5Ccong%20H_k(M%3B%5Cmathbb%7BZ%7D)%24%20which%20just%20means%20%24T_k%24%20is%20also%20trivial.%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%20Spring%20Semester%20Homework%203%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%23%20Problem%203%0A%0A%20%20%20%20**Using%20the%20cup%20product%20structure%2C%20show%20there%20is%20no%20map%20%24%5Cmathbb%7BR%7DP%5En%20%5Cto%20%5Cmathbb%7BR%7DP%5Em%24%20inducing%20a%20non-trivial%20map%20%24H%5E1(%5Cmathbb%7BR%7DP%5Em%3B%5Cmathbb%7BZ%7D_2)%24%20%24%5Cto%20H%5E1(%5Cmathbb%7BR%7DP%5En%3B%5Cmathbb%7BZ%7D_2)%24%20if%20%24n%20%3E%20m%24.%20What%20is%20the%20corresponding%20result%20for%20maps%20%24%5Cmathbb%7BC%7DP%5En%20%5Cto%20%5Cmathbb%7BC%7DP%5Em%24**%3F%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Suppose%20we%20have%20some%20%24f%3A%20%5Cmathbb%7BR%7DP%5En%20%5Cto%20%5Cmathbb%7BR%7DP%5Em%24%20which%20induces%20homomorphism%20%24f%5E*%3A%20H%5E1(%5Cmathbb%7BR%7DP%5Em%3B%20%5Cmathbb%7BZ%7D_2)%20%5Cto%20H%5E1(%5Cmathbb%7BR%7DP%5En%3B%5Cmathbb%7BZ%7D_2)%24.%20Since%20both%20groups%20are%20isomorphic%20to%20%24%5Cmathbb%7BZ%7D_2%24%2C%20the%20only%20non-trivial%20map%20is%20the%20map%20sending%20%24%5Calpha%24%20(generator%20of%20the%20domain)%20to%20%24%5Cbeta%24%20(generator%20of%20the%20codomain).%0A%0A%20%20%20%20Now%20consider%20the%20induced%20map%20between%20cohomology%20ring%20%24f%5E*%3A%20%5Cmathbb%7BZ%7D_2%5B%5Calpha%5D%2F(%5Calpha%5E%7Bm%2B1%7D)%20%5Cto%20%5Cmathbb%7BZ%7D_2%5B%5Cbeta%5D%2F(%5Cbeta%5E%7Bn%2B1%7D)%24%20where%20the%20former%20ring%20is%20generated%20by%20%24%5Calpha%24%20and%20the%20latter%20ring%20is%20generated%20by%20%24%5Cbeta%24%2C%20then%20by%20above%2C%20%24%5Calpha%24%20is%20sent%20to%20%24%5Cbeta%24.%20In%20the%20(domain)%20polynomial%20ring%20%24%5Calpha%5E%7Bm%2B1%7D%20%3D%200%24%2C%20so%20we%20have%20%240%20%3D%20f%5E*(%5Calpha%5E%7Bm%2B1%7D)%24%20%24%3D(f%5E*(%5Calpha))%5E%7Bm%2B1%7D%24%20%24%3D%5Cbeta%5E%7Bm%2B1%7D%24%20but%20it%20is%20not%20%240%24%20because%20%24n%2B1%3Em%2B1%24%2C%20so%20we%20have%20a%20contradiction.%0A%0A%20%20%20%20There%20is%20no%20map%20%24%5Cmathbb%7BC%7DP%5En%20%5Cto%20%5Cmathbb%7BC%7DP%5Em%24%20inducing%20a%20non-trivial%20map%20%24H%5E2(%5Cmathbb%7BC%7DP%5Em%3B%5Cmathbb%7BZ%7D)%24%20%24%5Cto%20H%5E2(%5Cmathbb%7BC%7DP%5En%3B%5Cmathbb%7BZ%7D)%24%3A%20we%20know%20these%20groups%20are%20(isomorphic%20to)%20%24%5Cmathbb%7BZ%7D%24%2C%20so%20the%20possible%20non-trivial%20maps%20map%20%24%5Calpha%24%20(generator%20of%20domain)%20to%20%24k%5Cbeta%24%20(%24%5Cbeta%24%20is%20generator%20of%20codomain%2C%20%24k%24%20is%20non-zero%20integer).%20Again%20consider%20the%20induced%20map%20between%20cohomology%20ring%20%24f%5E*%3A%20%5Cmathbb%7BZ%7D%5B%5Calpha%5D%2F(%5Calpha%5E%7Bm%2B1%7D)%20%5Cto%20%5Cmathbb%7BZ%7D%5B%5Cbeta%5D%2F(%5Cbeta%5E%7Bn%2B1%7D)%24%20that%20maps%20%24%5Calpha%24%20to%20%24k%20%5Cbeta%24%2C%20it%20maps%20%240%20%3D%20a%5E%7Bm%2B1%7D%24%20to%20%24(k%5Cbeta)%5E%7Bm%2B1%7D%24%20which%20is%20not%20zero%20by%20the%20same%20reason%2C%20and%20we%20have%20a%20contradiction%20again.%0A%0A%20%20%20%20%23%23%20Problem%207%0A%0A%20%20%20%20**Use%20cup%20products%20to%20show%20that%20%24%5Cmathbb%7BR%7DP%5E3%24%20is%20not%20homotopy%20equivalent%20to%20%24%5Cmathbb%7BR%7DP%5E2%20%5Cvee%20S%5E3%24**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20The%20cohomology%20ring%20of%20%24%5Cmathbb%7BR%7DP%5E3%24%20with%20%24%5Cmathbb%7BZ%7D_2%24%20coefficients%20is%20%24%5Cmathbb%7BZ%7D_2%5B%5Calpha%5D%2F(%5Calpha%5E4)%24%20(generated%20by%20%241%2C%20%5Calpha%2C%20%5Calpha%5E2%2C%20%5Calpha%5E3%24%20in%20each%20dimension%2C%20respectively).%20If%20we%20consider%20the%20reduced%20version%20then%20%24%5Ctilde%7BH%5E0%7D(%5Cmathbb%7BR%7DP%5E3%3B%5Cmathbb%7BZ%7D_2)%20%3D%200%24%20(the%20%240%24-dimensional%20generator%20disappears)%20thus%20%24%5Ctilde%7BH%5E*%7D(%5Cmathbb%7BR%7DP%5E3%3B%5Cmathbb%7BZ%7D_2)%24%20is%20the%20sub-ring%20%24R_1%24%20of%20%24%5Cmathbb%7BZ%7D_2%5B%5Calpha%5D%2F(%5Calpha%5E4)%24%20consisting%20of%20polynomials%20without%20constant%20term.%0A%0A%20%20%20%20Similarly%2C%20%24%5Ctilde%7BH%5E*%7D(%5Cmathbb%7BR%7DP%5E2%3B%5Cmathbb%7BZ%7D_2)%24%20is%20the%20sub-ring%20%24R_2%24%20of%20%24%5Cmathbb%7BZ%7D_2%5B%5Cbeta%5D%2F(%5Cbeta%5E3)%24%20without%20constant%20term%2C%20%24%5Ctilde%7BH%5E*%7D(S%5E3%3B%5Cmathbb%7BZ%7D_2)%24%20is%20the%20sub-ring%20%24R_3%24%20of%20%24%5Cmathbb%7BZ%7D_2%5B%5Cgamma%5D%2F(%5Cgamma%5E2)%24%20without%20constant%20term.%20Since%20the%20reduced%20cohomology%20ring%20of%20wedge%20sum%20is%20just%20product%20of%20reduced%20cohomology%20rings%2C%20we%20get%20that%20%24%5Ctilde%7BH%5E*%7D(%5Cmathbb%7BR%7DP%5E2%20%5Cvee%20S%5E3%3B%5Cmathbb%7BZ%7D_2)%20%5Ccong%20R_2%20%5Ctimes%20R_3%24.%0A%0A%20%20%20%20But%20now%20%24R_1%24%20and%20%24R_2%20%5Ctimes%20R_3%24%20are%20not%20isomorphic%3A%20%24%5Calpha%20%5Cin%20R_1%24%20is%20an%20element%20with%20order%20%244%24%20but%20there%20is%20no%20such%20element%20in%20%24R_2%20%5Ctimes%20R_3%24.%0A%0A%20%20%20%20%23%23%20Problem%2011%0A%0A%20%20%20%20**Using%20cup%20products%2C%20show%20that%20every%20map%20%24S%5E%7Bk%2Bl%7D%20%5Cto%20S%5Ek%20%5Ctimes%20S%5El%24%20induces%20the%20trivial%20homomorphism%20%24H_%7Bk%2Bl%7D(S%5E%7Bk%2Bl%7D)%20%5Cto%20H_%7Bk%2Bl%7D(S%5Ek%20%5Ctimes%20S%5El)%24%2C%20assuming%20%24k%3E0%24%20and%20%24l%3E0%24**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Let%20%24f%24%20be%20a%20(continuous)%20map%20%24S%5E%7Bk%2Bl%7D%20%5Cto%20S%5Ek%20%5Ctimes%20S%5El%24%2C%20%24p_1%24%20and%20%24p_2%24%20are%20projection%20from%20%24S%5Ek%20%5Ctimes%20S%5El%24%20to%20%24S%5Ek%24%20and%20%24S%5El%24%2C%20respectively.%20Then%20%24f%24%20induces%20homomorphism%20%24H%5E%7Bk%2Bl%7D(S%5Ek%20%5Ctimes%20S%5El)%20%5Cto%20H%5E%7Bk%2Bl%7D(S%5E%7Bk%2Bl%7D)%24.%20By%20the%20K%C3%BCnneth%20formula%2C%20the%20domain%20is%20isomorphic%20to%20%24%5Cbigoplus%5Climits_%7Bi%2Bj%3Dk%2Bl%7DH%5Ei(S%5Ek)%20%5Cotimes_%7B%5Cmathbb%7BZ%7D%7D%20H%5Ej(S%5El)%24%20%24%5Ccong%20H%5Ek(S%5Ek)%20%5Cotimes_%7B%5Cmathbb%7BZ%7D%7D%20H%5El(S%5El)%24%20%24%5Ccong%20%5Cmathbb%7BZ%7D%20%5Cotimes_%7B%5Cmathbb%7BZ%7D%7D%20%5Cmathbb%7BZ%7D%24%20%24%5Ccong%20%5Cmathbb%7BZ%7D%24%20is%20generated%20by%20%24a%20%5Cotimes%20b%24%20where%20%24a%2C%20b%24%20are%20generators%20for%20%24H%5Ek(S%5Ek)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%20and%20%24H%5El(S%5El)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%2C%20respectively.%20The%20isomorphism%20(in%20the%20K%C3%BCnneth%20formula)%20is%20realized%20by%20%24a%20%5Cotimes%20b%20%5Cleftrightarrow%20a%5Ctimes%20b%24%20%24%3Dp_1%5E*(a)%20%5Ccup%20p_2%5E*(b)%24.%0A%0A%20%20%20%20Write%20%24f%5E*(a%20%5Ctimes%20b)%24%20%24%3D%20f%5E*(p_1%5E*(a)%20%5Ccup%20p_2%5E*(b))%24%20%24%3Df%5E*(p_1%5E*(a))%20%5Ccup%20f%5E*(p_2%5E*(b))%24%20%24%3D((p_1%5Ccirc%20f)%5E*(a))%5Ccup((p_2%5Ccirc%20f)%5E*(b))%24%20%24%3D%200%20%5Ccup%200%20%3D%200%24%2C%20because%20%24(p_1%5Ccirc%20f)%5E*%3A%20H%5Ek(S%5Ek)%20%5Cto%20H%5Ek(S%5E%7Bk%2Bl%7D)%3D0%24%20and%20similar%20for%20%24p_2%24%2C%20thus%20%24f%5E*%24%20maps%20the%20only%20generator%20to%20%240%24%20thus%20is%20the%20trivial%20homomorphism%20between%20cohomology%2C%20thus%20its%20dual%20%24f_*%24%20is%20also%20the%20trivial%20homomorphism%20between%20homology.%0A%0A%20%20%20%20%23%23%20Problem%2032%0A%0A%20%20%20%20**Show%20that%20a%20compact%20manifold%20does%20not%20retract%20onto%20its%20boundary**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20For%20simplicity%20of%20notation%2C%20we%20assume%20to%20use%20%24%5Cmathbb%7BZ%7D_2%24%20coefficients%20in%20all%20cohomology%2Fhomology%20groups%20in%20this%20problem.%0A%0A%20%20%20%20Suppose%20we%20have%20a%20compact%20%24n%24-manifold%20%24M%24%20with%20boundary%20%24%5Cpartial%20M%24%2C%20and%20suppose%20%24r%3A%20M%20%5Cto%20%5Cpartial%20M%24%20is%20a%20retraction%20(%24r%20%5Ccirc%20i%20%3D%20%5Ctext%7Bid%7D_%7B%5Cpartial%20M%7D%24).%20We%20can%20further%20assume%20%24M%24%20is%20connected%20because%20the%20retraction%20should%20take%20place%20component-wise%20anyway.%20Consider%20the%20LES%20of%20%24(M%2C%20%5Cpartial%20M)%24%3A%20%24%24%5Cbegin%7Barray%7D%7Bcccccccc%7D%20%5Cto%20%26%20H_n(%5Cpartial%20M)%20%26%20%5Cto%20%26%20H_n(M)%20%26%20%5Cto%20%26%20H_n(M%2C%20%5Cpartial%20M)%20%26%20%5Cto%20%26%20%5C%5C%20%5Cto%20%26%20H_%7Bn-1%7D(%5Cpartial%20M)%20%26%20%5Cxrightarrow%7Bi_*%7D%20%26%20H_%7Bn-1%7D(M)%20%26%20%5Cto%20%26%20H_%7Bn-1%7D(M%2C%20%5Cpartial%20M)%20%26%20%5Cto%20%26%20%5Cend%7Barray%7D%24%24%20We%20have%20that%3A%0A%0A%20%20%20%201.%20By%20retraction%2C%20the%20induced%20homomorphism%20%24i_*%24%20is%20injective%3B%0A%20%20%20%202.%20Since%20all%20manifolds%20are%20%24%5Cmathbb%7BZ%7D_2%24-orientable%2C%20Poincar%C3%A9-Lefschetz%20Duality%20applies.%20%24H_n(M)%20%5Ccong%20H%5E0(M%2C%5Cpartial%20M)%24%20which%20%24%5Ccong%20%5Ctilde%7BH%5E0%7D(M%2F%5Cpartial%20M)%24%20which%20is%20%240%24%20because%20%24M%24%20is%20assumed%20to%20be%20connected%3B%0A%20%20%20%203.%20By%20orientability%2C%20%24H_n(M%2C%5Cpartial%20M)%20%5Ccong%20%5Cmathbb%7BZ%7D_2%24%3B%0A%0A%20%20%20%20So%20the%20following%20part%3A%20%24%24H_n(M)%20%5Cto%20H_n(M%2C%5Cpartial%20M)%20%5Cto%20H_%7Bn-1%7D(%5Cpartial%20M)%20%5Cto%20H_%7Bn-1%7D(M)%24%24%20is%20the%20same%20as%3A%20%24%240%20%5Cto%5E0%20%5Cmathbb%7BZ%7D_2%20%5Cxrightarrow%7B%5Cpartial%7D%20H_%7Bn-1%7D(%5Cpartial%20M)%20%5Cxrightarrow%7Bi_*%7D%20H_%7Bn-1%7D(M).%24%24%20In%20particular%2C%20%24%5Cker%7B%5Cpartial%7D%20%3D%20%5Ctext%7Bim%7D(0)%20%3D%200%24%2C%20thus%20%24%5Ctext%7Bim%7D(%5Cpartial)%20%3D%20%5Cmathbb%7BZ%7D_2%24%20thus%20%24%5Cker%7Bi_*%7D%20%3D%20%5Cmathbb%7BZ%7D_2%24%2C%20thus%20we%20have%20a%20contradiction%20because%20%24i_*%24%20should%20be%20injective.%0A%0A%20%20%20%20%23%23%20Problem%2033%0A%0A%20%20%20%20**Show%20that%20if%20%24M%24%20is%20a%20compact%20contractible%20%24n%24-manifold%20then%20%24%5Cpartial%20M%24%20is%20a%20homology%20%24(n-1)%24-sphere%2C%20that%20is%2C%20%24H_i(%5Cpartial%20M%3B%5Cmathbb%7BZ%7D)%20%5Ccong%20H_i(S%5E%7Bn-1%7D%3B%5Cmathbb%7BZ%7D)%24%20for%20all%20%24i%24**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20A%20point%20does%20not%20have%20boundary%3B%20a%20compact%20contractible%20%241%24-manifold%20can%20only%20be%20a%20line%20segment%2C%20in%20which%20case%20%24%5Cpartial%20M%24%20equals%20two%20points%20and%20is%20a%20(homology)%20%240%24-sphere.%20So%20we%20can%20assume%20%24n%5Cge%202%24.%0A%0A%20%20%20%20Suppose%20%24M%24%20is%20orientable%2C%20then%20for%20the%20following%20part%20of%20LES%20of%20pair%20%24%24H_n(M)%20%5Cto%20H_n(M%2C%5Cpartial%20M)%20%5Cto%20H_%7Bn-1%7D(%5Cpartial%20M)%20%5Cto%20H_%7Bn-1%7D(M)%2C%24%24%20we%20have%20%24H_n(M)%2C%20H_%7Bn-1%7D(M)%20%5Ccong%200%24%20(contractible)%2C%20thus%20%24H_%7Bn-1%7D(%5Cpartial%20M)%24%20%24%5Ccong%20H_n(M%2C%20%5Cpartial%20M)%24%20%24%5Ccong%20%5Cmathbb%7BZ%7D%24%20by%20definition%20of%20orientability.%20For%20other%20%24i%20%5Cne%200%24%2C%20use%20%24H_i(%5Cpartial%20M)%24%20%24%5Ccong%20H_%7Bi%2B1%7D(M%2C%5Cpartial%20M)%24%20%24%5Ccong%5E%7B%5Ctext%7BPL%20Duality%7D%7D%20H%5E%7Bn-i-1%7D(M)%24%20%24%5Ccong%200%24%20because%20%24M%24%20is%20contractible%2C%20and%20use%20reduced%20(co)homology%20for%20%24i%20%3D%200%24%20to%20get%20%24%5Ctilde%7BH%7D_0(%5Cpartial%20M)%20%3D%200%24%2C%20we%20get%20that%20%24H_i(%5Cpartial%20M)%20%5Ccong%20H_i(S%5E%7Bn-1%7D)%24%20for%20all%20%24i%24.%0A%0A%20%20%20%20In%20fact%2C%20%24M%24%20can%20only%20be%20orientable.%20By%20Prop%203.42%2C%20%24%5Cpartial%20M%24%20has%20a%20collar%20neighborhood%20%24c(M)%5Ccong%20%5Cpartial%20M%20%5Ctimes%20%5B0%2C1)%24.%20Consider%20%24M%5Csetminus%20c(M)%24%2C%20then%20%24M%24%20deformation%20retracts%20onto%20it%20by%20%24H(x%2Ct)%24%20where%20%24H(x%2Ct)%3Dx%24%20if%20%24x%5Cin%20M%5Csetminus%20c(M)%24%20and%20%24H(x%2Ct)%24%20equals%20the%20homeomorphic%20image%20of%20%24x%24%20under%20%24c(M)%20%5Ccong%20%5Cpartial%20M%20%5Ctimes%20%5B1-t%2C1)%24%20if%20%24x%20%5Cin%20c(M)%24.%20%24%5Ctext%7Bint%7D(M)%24%20also%20deformation%20retracts%20onto%20it%20(similar%20%24H%24%2C%20just%20use%20the%20homeomorphism%20%24c(M)%20%5Csetminus%20%5Cpartial%20M%20%5Ccong%20%5Cpartial%20M%5Ctimes(0%2C1)%24%20instead).%0A%0A%20%20%20%20(Note%3A%20interior%20of%20contractible%20space%20is%20not%20necessarily%20contractible%2C%20so%20we%20are%20not%20'making%20things%20complicated'%20above)%0A%0A%20%20%20%20Thus%20%24%5Ctext%7Bint%7D(M)%3D%20M%5Csetminus%20%5Cpartial%20M%24%20is%20homotopy%20equivalent%20with%20%24M%24%20thus%20contractible%2C%20thus%20simply%20connected%2C%20thus%20orientable%2C%20thus%20%24M%24%20is%20orientable%20(p.%20253).%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%20Spring%20Semester%20Homework%204%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%23%20Problem%203%0A%0A%20%20%20%20**For%20a%20path-connected%20space%20%24X%24%2C%20show%20that%20%24%5Cpi_1(X)%24%20is%20abelian%20if%20and%20only%20if%20all%20base-point-change%20homomorphisms%20%24%5Cbeta_h%24%20depend%20only%20on%20the%20endpoints%20of%20the%20path%20%24h%24**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Suppose%20%24%5Cpi_1(X%2Cx_0)%24%20is%20abelian%2C%20and%20say%20%24g%2C%20h%24%20are%20two%20paths%20from%20%24x_1%24%20to%20%24x_0%24.%20Let%20%24f%24%20be%20an%20arbitrary%20loop%20based%20at%20%24x_0%24%2C%20we%20want%20to%20show%20%24%5Cbeta_g(%5Bf%5D)%20%3D%20%5Cbeta_h(%5Bf%5D)%24.%20We%20have%3A%20%24%24%5Cbegin%7Baligned%7D%5Cbeta_g(%5Bf%5D)%20%26%3A%3D%20%5Bg%5E%7B-1%7D%5Ccdot%20f%5Ccdot%20g%5D%20%5C%5C%20%26%3D%5Bh%5E%7B-1%7D%5Ccdot%20h%5Ccdot%20g%5E%7B-1%7D%20%5Ccdot%20f%20%5Ccdot%20g%20%5Ccdot%20h%5E%7B-1%7D%20%5Ccdot%20h%5D%20%5C%5C%20%26%3D%5Bh%5E%7B-1%7D%5Ccdot%20(h%5Ccdot%20g%5E%7B-1%7D)%5Ccdot%20f%20%5Ccdot%20(h%5Ccdot%20g%5E%7B-1%7D)%5E%7B-1%7D%5Ccdot%20h%5D%5C%5C%26%3D%3A%5Cbeta_h(%5B(h%5Ccdot%20g%5E%7B-1%7D)%5Ccdot%20f%20%5Ccdot%20(h%5Ccdot%20g%5E%7B-1%7D)%5E%7B-1%7D%5D)%5C%5C%26%3D%5Cbeta_h(%5Bh%5Ccdot%20g%5E%7B-1%7D%5D%20%5Bf%5D%20%5B(h%5Ccdot%20g%5E%7B-1%7D)%5E%7B-1%7D%5D)%5C%5C%26%3D%5Cbeta_h(%5Bf%5D).%5Cend%7Baligned%7D%24%24%20So%20we%20get%20the%20desired%20result.%20The%20last%20step%20is%20due%20to%20that%20%24%5Bh%5Ccdot%20g%5E%7B-1%7D%5D%2C%5Bf%5D%2C%5B(h%5Ccdot%20g%5E%7B-1%7D)%5E%7B-1%7D%5D%24%20are%20all%20elements%20in%20%24%5Cpi_1(X%2Cx_0)%24%20which%20is%20assumed%20to%20be%20abelian.%0A%0A%20%20%20%20Suppose%20then%20%24%5Cpi_1(X%2C%20x_0)%24%20is%20not%20abelian%2C%20then%20we%20can%20find%20%24f%2C%20g%24%20loops%20based%20at%20%24x_0%24%20such%20that%20%24%5Bf%5D%5Bg%5D%20%5Cne%20%5Bg%5D%5Bf%5D%24.%20Denote%20%24%5Cmathbb%7B1%7D_%7Bx_0%7D%24%20as%20the%20identity%20(constant%20loop%20at%20%24x_0%24)%2C%20we%20have%20that%20%24%5Cbeta_g(%5Bf%5D)%3A%20%3D%20%5Bg%5E%7B-1%7D%5Ccdot%20f%20%5Ccdot%20g%5D%24%20%24%5Cne%20%5Bf%5D%20%3D%3A%20%5Cbeta_%7B%5Cmathbb%7B1%7D_%7Bx_0%7D%7D(%5Bf%5D)%24%20(inequality%20is%20because%20inverse%20is%20unique%20in%20a%20group)%2C%20so%20we%20find%20two%20paths%20%24g%2C%20%5Cmathbb%7B1%7D_%7Bx_0%7D%24%20with%20the%20same%20endpoint%20but%20different%20induced%20homomorphism%2C%20and%20the%20statement%20is%20proved.%0A%0A%20%20%20%20%23%23%20Problem%205%0A%0A%20%20%20%20**Show%20that%20for%20a%20space%20%24X%24%2C%20the%20following%20three%20conditions%20are%20equivalent**%3A%0A%0A%20%20%20%201.%20**Every%20map%20%24S%5E1%20%5Cto%20X%24%20is%20homotopic%20to%20a%20constant%20map%2C%20with%20image%20a%20point**%3B%0A%20%20%20%202.%20**Every%20map%20%24S%5E1%20%5Cto%20X%24%20extends%20to%20a%20map%20%24D%5E2%20%5Cto%20X%24**%3B%0A%20%20%20%203.%20**%24%5Cpi_1(X%2C%20x_0)%20%3D%200%24%20for%20all%20%24x_0%20%5Cin%20X%24**.%0A%0A%20%20%20%20**Deduce%20that%20a%20space%20%24X%24%20is%20simply%20connected%20if%20and%20only%20if%20all%20maps%20%24S%5E1%20%5Cto%20X%24%20are%20homotopic**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20(%241%20%5Cimplies%202%24)%20Suppose%20we%20have%20a%20map%20%24f%3A%20S%5E1%20%5Cto%20X%24.%20By%20assumption%20we%20have%20%24H(x%2Ct)%24%20such%20that%20%24h_1(x)%20%3D%20f(x)%24%20and%20%24h_0(x)%20%3D%20x_0%24%20for%20some%20fixed%20point%20%24x_0%20%5Cin%20X%24%20and%20%24x%20%5Cin%20S%5E1%24.%20Notice%20every%20point%20in%20%24D%5E2%24%20can%20be%20expressed%20as%20%24tx%24%20where%20%24x%20%5Cin%20S%5E1%24%20and%20%24t%20%5Cin%20%5B0%2C1%5D%24%2C%20and%20the%20pair%20%24x%2C%20t%24%20is%20unique%20unless%20%24t%20%3D%200%24.%0A%0A%20%20%20%20Define%20%24%5Ctilde%7Bf%7D%3A%20D%5E2%20%5Cto%20X%24%20by%20%24%5Ctilde%7Bf%7D(tx)%20%3D%20H(x%2Ct)%24%2C%20then%20%24%5Ctilde%7Bf%7D%24%20is%20well-defined%20because%20%24H(x%2C0)%20%3D%20x_0%24%20for%20all%20%24x%24%2C%20%24%5Ctilde%7Bf%7D%7C_%7BS%5E1%7D%20%3D%20h_1%20%3D%20f%24%2C%20and%20%24%5Ctilde%7Bf%7D%24%20is%20continuous.%20So%20we%20found%20a%20desired%20extension.%0A%0A%20%20%20%20(%242%20%5Cimplies%203%24)%20Let%20%24f%24%20be%20an%20arbitrary%20loop%20based%20at%20%24x_0%24%2C%20which%20can%20be%20understand%20as%20a%20function%20%24f%3A(S%5E1%2C1)%20%5Cto%20(X%2Cx_0)%24.%20We%20need%20to%20show%20that%20%24%5Bf%5D%20%3D%20%5Cmathbb%7B1%7D_%7Bx_0%7D%24.%20By%20%242%24%2C%20we%20can%20get%20an%20extension%20%24%5Ctilde%7Bf%7D%3A%20D%5E2%20%5Cto%20X%24.%20In%20particular%2C%20%24f%20%3D%20%5Ctilde%7Bf%7D%20%5Ccirc%20i%24%20where%20%24i%3A%20S%5E1%20%5Cto%20D%5E2%24%20is%20inclusion.%20Then%20by%20property%20of%20the%20induced%20homomorphisms%2C%20we%20get%20%24f_*%20%3D%20(%5Ctilde%7Bf%7D%20%5Ccirc%20i)_*%24%20%24%3D%20%5Ctilde%7Bf%7D_*%20%5Ccirc%20i_*%24%20%24%3A%20%5Cpi_1(S_1%2C%201)%20%5Cto%20%5Cpi_1(D%5E2%2C%201)%20%5Cto%20%5Cpi_1(X%2Cx_0)%24.%20Since%20%24D%5E2%24%20is%20simply%20connected%2C%20%24%5Cpi_1(D%5E2%2C1)%24%20is%20trivial%2C%20thus%20%24i_*%24%20thus%20the%20composition%20%24f_*%3A%20%5Cpi_1(S%5E1%2C%201)%20%5Cto%20%5Cpi_1(X%2Cx_0)%24%20is%20trivial%2C%20which%20means%20that%20%24f_*(%5B%5Cgamma%5D)%20%3D%20%5Bf%5Ccirc%20%5Cgamma%5D%20%3D%20%5Cmathbb%7B1%7D_%7Bx_0%7D%24%20for%20any%20loop%20%24%5Cgamma%24%20in%20%24S%5E1%24%20based%20at%20%241%24.%20Take%20%24%5Cgamma%24%20to%20be%20the%20map%20that%20goes%20%24S%5E1%24%20once%2C%20we%20get%20%24%5Bf%5D%20%3D%20%5Bf%5Ccirc%20%5Cgamma%5D%20%3D%20%5Cmathbb%7B1%7D_%7Bx_0%7D%24%2C%20as%20desired.%0A%0A%20%20%20%20(%243%20%5Cimplies%201%24)%20This%20is%20essentially%20by%20definition%20of%20fundamental%20group.%0A%0A%20%20%20%20By%20definition%2C%20%24X%24%20being%20simply%20connected%20if%20and%20only%20if%20%24X%24%20is%20path%20connected%20and%20has%20trivial%20fundamental%20group%20(%243%24)%2C%20which%20we%20proved%20is%20equivalent%20with%20that%20%24X%24%20is%20path%20connected%20and%20all%20maps%20%24S%5E1%20%5Cto%20X%24%20is%20homotopic%20to%20a%20constant%20map%20(%241%24)%20-%20which%20is%20not%20hard%20to%20be%20seen%20equivalent%20with%20that%20all%20maps%20%24S%5E1%20%5Cto%20X%24%20are%20homotopic%20(a%20constant%20map%20%24S%5E1%20%5Cto%20X%24%20is%20a%20map%20%24S%5E1%20%5Cto%20X%24%20in%20the%20first%20place%2C%20and%20homotopy%20is%20an%20equivalent%20relation).%0A%0A%20%20%20%20%23%23%20Problem%206%0A%0A%20%20%20%20**We%20can%20regard%20%24%5Cpi_1(X%2Cx_0)%24%20as%20the%20set%20of%20base-point-preserving%20homotopy%20classes%20of%20maps%20%24(S%5E1%2Cs_0)%20%5Cto%20(X%2C%20x_0)%24.%20Let%20%24%5BS%5E1%2CX%5D%24%20be%20the%20set%20of%20homotopy%20classes%20of%20maps%20%24S%5E1%20%5Cto%20X%24%2C%20with%20no%20conditions%20on%20base-points.%20Thus%20there%20is%20a%20natural%20map%20%24%5CPhi%3A%20%5Cpi_1(X%2C%20x_0)%20%5Cto%20%5BS%5E1%2C%20X%5D%24%20obtained%20by%20ignoring%20base-points.%20Show%20that%20%24%5CPhi%24%20is%20onto%20if%20%24X%24%20is%20path-connected%2C%20and%20that%20%24%5CPhi(%5Bf%5D)%20%3D%20%5CPhi(%5Bg%5D)%24%20if%20and%20only%20if%20%24%5Bf%5D%24%20and%20%24%5Bg%5D%24%20are%20conjugate%20in%20%24%5Cpi_1(X%2C%20x_0)%24.%20Hence%20%24%5CPhi%24%20induces%20a%20one-to-one%20correspondence%20between%20%24%5BS%5E1%2CX%5D%24%20and%20the%20set%20of%20conjugacy%20classes%20in%20%24%5Cpi_1(X)%24%20when%20%24X%24%20is%20path-connected**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20First%20we%20show%20%24%5CPhi%24%20is%20onto.%20That%20is%2C%20for%20any%20%24f%3A%20S%5E1%20%5Cto%20X%24%2C%20we%20would%20like%20to%20find%20some%20element%20in%20%24%5Cpi(X%2Cx_0)%24%20so%20that%20its%20image%20under%20%24%5CPhi%24%20is%20the%20homotopy%20class%20%24%5Bf%5D%24.%0A%0A%20%20%20%20Observe%20that%20if%20%24%5Calpha%24%20is%20a%20loop%20based%20at%20%24a%24%20and%20%24%5Cbeta%24%20is%20a%20path%20connecting%20%24a%24%20and%20%24b%24%2C%20then%20ignoring%20the%20base%20point%2C%20%24%5Cbeta%5E%7B-1%7D%5Ccdot%20%5Calpha%20%5Ccdot%20%5Cbeta%24%20is%20homotopic%20with%20%24%5Calpha%20%5Ccdot%20%5Cbeta%20%5Ccdot%20%5Cbeta%5E%7B-1%7D%20%3D%20%5Calpha%24%20(we%20can%20think%20of%20it%20as%20we%20rotate%20the%20path%20along%20the%20path%2C%20the%20base%20point%20changes%20from%20%24a%24%20to%20%24b%24).%20Now%2C%20understand%20%24f%24%20as%20a%20loop%20based%20at%20some%20%24x_1%20%5Cin%20X%24.%20Since%20%24X%24%20is%20path-connected%2C%20we%20can%20find%20a%20path%20%24g%3AI%20%5Cto%20X%24%20with%20%24g(0)%20%3D%20x_1%2C%20g(1)%20%3D%20x_0%24%2C%20so%20that%20%24g%5E%7B-1%7D%20%5Ccdot%20f%20%5Ccdot%20g%24%20is%20a%20loop%20based%20at%20%24x_0%24.%20By%20the%20above%20observation%20and%20the%20definition%20of%20%24%5CPhi%24%2C%20we%20have%20that%20%24%5CPhi(%5Bg%5E%7B-1%7D%20%5Ccdot%20f%20%5Ccdot%20g%5D)%24%20%24%3D%5CPhi(%5Bf%5D)%20%3D%20%5Bf%5D%24%2C%20so%20we%20have%20found%20the%20pre-image.%0A%0A%20%20%20%20The%20above%20equation%20%24%5CPhi(%5Bg%5E%7B-1%7D%5Ccdot%20f%20%5Ccdot%20g%5D)%20%3D%20%5CPhi(%5Bf%5D)%24%20also%20proves%20that%20if%20%24%5Bf%5D%24%20and%20%24%5Bh%5D%20%3D%20%5Bg%5E%7B-1%7D%5D%5Bf%5D%5Bg%5D%24%20for%20certain%20%24%5Bg%5D%24%20are%20conjugate%2C%20then%20their%20image%20under%20%24%5CPhi%24%20is%20the%20same.%0A%0A%20%20%20%20Now%2C%20suppose%20%24f%24%20and%20%24g%24%20are%20loops%20based%20at%20%24x_0%24%20and%20%24%5CPhi(%5Bf%5D)%20%3D%20%5CPhi(%5Bg%5D)%24%2C%20we%20need%20to%20show%20%24%5Bf%5D%24%20and%20%24%5Bg%5D%24%20are%20conjugate%20in%20the%20fundamental%20group.%0A%0A%20%20%20%20By%20the%20definition%20of%20%24%5CPhi%24%2C%20the%20assumption%20means%20that%20%24f%24%20and%20%24g%24%20are%20homotopic%20(ignoring%20base%20point)%2C%20i.e.%20there%20is%20%24H(x%2Ct)%3AS%5E1%20%5Ctimes%20I%20%5Cto%20X%24%20with%20%24h_0%20%3D%20f%2C%20h_1%20%3D%20g%24.%20We%20have%20that%20%24H(1%2C0)%24%20(base%20point%20of%20%24f%24)%20and%20%24H(1%2C%201)%24%20(base%20point%20of%20%24g%24)%20are%20both%20%24x_0%24%2C%20but%20%24H(1%2C%20t)%2C%200%3Ct%3C1%24%20may%20not%20be%20this%20point.%20Define%20%24h%3A%20I%20%5Cto%20X%24%20by%20%24h(t)%20%3D%20H(1%2C%20t)%24%2C%20then%20%24h%24%20is%20continuous%2C%20%24h(0)%20%3D%20h(1)%20%3D%20x_0%24%20so%20it%20is%20a%20loop%20based%20at%20%24x_0%24.%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230223092327.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20Then%20%24f%24%20and%20%24h%5Ccdot%20g%20%5Ccdot%20h%5E%7B-1%7D%24%20are%20homotopic%20with%20base%20point%20fixed%2C%20realized%20by%20the%20homotopy%20%24L(x%2Ct)%24%20where%20%24l_0%20%3D%20f%24%2C%20%24l_1%20%3D%20h%5Ccdot%20g%20%5Ccdot%20h%5E%7B-1%7D%24%2C%20and%20for%20%240%3Ct%3C1%24%2C%20%24l_t%24%20be%20the%20loop%20(based%20at%20%24x_0%24)%20%24h%7C_%7B%5B0%2Ct%5D%7D%20%5Ccdot%20h_t%20%5Ccdot%20h%7C_%7B%5B0%2Ct%5D%7D%5E%7B-1%7D%24%20(here%20%24h_t%24%20is%20the%20loop%20defined%20above%20in%20%24H(x%2Ct)%24%2C%20it%20is%20a%20loop%20based%20at%20%24H(1%2Ct)%20%3D%20h(t)%24).%20In%20particular%2C%20%24%5Bf%5D%20%3D%20%5Bh%5Ccdot%20g%20%5Ccdot%20h%5E%7B-1%7D%5D%24%20so%20%24%5Bf%5D%24%20and%20%24%5Bg%5D%24%20are%20conjugate%20by%20%24%5Bh%5D%24%20in%20the%20fundamental%20group.%0A%0A%20%20%20%20%23%23%20Problem%2016%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230217161525.png%22%20width%3D%22400%22%20%2F%3E%0A%0A%20%20%20%20**Show%20that%20there%20are%20no%20retractions%20%24r%3A%20X%20%5Cto%20A%24%20in%20the%20following%20cases**%3A%0A%0A%20%20%20%201.%20**%24X%20%3D%20%5Cmathbb%7BR%7D%5E3%24%20with%20%24A%24%20any%20subspace%20homeomorphic%20to%20%24S%5E1%24**%3B%0A%20%20%20%202.%20**%24X%20%3D%20S%5E1%20%5Ctimes%20D%5E2%24%20with%20%24A%24%20its%20boundary%20torus%20%24S%5E1%20%5Ctimes%20S%5E1%24**%3B%0A%20%20%20%203.%20**%24X%20%3D%20S%5E1%20%5Ctimes%20D%5E2%24%20with%20%24A%24%20the%20circle%20shown%20in%20the%20figure**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%201.%20%24%5Cpi_1(X)%20%3D%201%24%20and%20%24%5Cpi_1(A)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%2C%20so%20there%20is%20no%20injective%20homomorphism%20from%20%24%5Cpi_1(A)%20%5Cto%20%5Cpi_1(X)%24%3B%0A%20%20%20%202.%20%24%5Cpi_1(X)%20%5Ccong%20%5Cpi_1(S%5E1)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%20and%20%24%5Cpi_1(A)%20%5Ccong%20%5Cmathbb%7BZ%7D%5E2%24%2C%20so%20again%20there%20is%20no%20injective%20homomorphism%20from%20%24%5Cpi_1(A)%20%5Cto%20%5Cpi_1(X)%24%3B%0A%20%20%20%203.%20We%20know%20for%20a%20retract%20%24r%24%20we%20should%20have%20inclusion%20so%20that%20%24(r%20%5Ccirc%20i)_*%20%3D%20r_*%20%5Ccirc%20i_*%20%3D%20%5Ctext%7Bid%7D%24.%20By%20definition%2C%20%24i_*%3A%20%5Cpi_1(A)%20%5Cto%20%5Cpi_1(X)%24%20sends%20the%20generator%20of%20%24%5Cpi_1(A)%24%2C%20which%20is%20a%20loop%20goes%20around%20%24A%24%20once%20(in%20either%20direction)%2C%20to%20the%20same%20loop%20in%20%24X%24.%20But%20this%20loop%20in%20%24X%24%20contract%20to%20a%20constant%20map%20-%20because%20now%20we%20can%20use%20any%20point%20in%20this%20solid%20torus%2C%20also%20there%20is%20no%20reason%20to%20not%20be%20able%20to%20self-cross.%20So%20%24i_*%24%20is%20trivial%2C%20thus%20%24(r%5Ccirc%20i)_*%24%20cannot%20be%20identity.%0A%0A%20%20%20%20%23%23%20Problem%203%0A%0A%20%20%20%20**Show%20that%20the%20complement%20of%20a%20finite%20set%20of%20points%20in%20%24%5Cmathbb%7BR%7D%5En%24%20is%20simply%20connected%20if%20%24n%20%5Cge%203%24**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Say%20the%20finite%20set%20of%20points%20are%20%24X%20%3D%20%5Clbrace%20x_1%2C%20%5Cdots%2C%20x_m%20%5Crbrace%24%20so%20the%20space%20we%20are%20interested%20at%20is%20%24%5Cmathbb%7BR%7D%5En%20%5Csetminus%20X%24.%20Deformation%20retract%20%24%5Cmathbb%7BR%7D%5En%20%5Csetminus%20X%24%20to%20the%20space%20%24%5Cmathbb%7BR%7D%5En%24%20with%20%24(1%2C0%2C%5Cdots%2C0)%2C%5Cdots%2C(m%2C0%2C%5Cdots%2C0)%24%20removed.%20For%20simplicity%20we%20illustrate%20it%20in%20%24%5Cmathbb%7BR%7D%5E2%24%2C%20but%20the%20idea%20is%20the%20same%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230224100139.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20This%20space%20deformation%20retract%20to%20a%20space%20of%20wedge%20sum%20of%20%24m%24%20copies%20of%20%24(n-1)%24-pheres%20of%20radius%20%24%5Cfrac%7B1%7D%7B2%7D%24%2C%20and%20we%20can%20further%20deformation%20retract%20this%20space%20to%20%24m%24%20copies%20of%20%24(n-1)%24-spheres%20with%20a%20single%20connected%20point.%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230224101137.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20This%20space%20is%20simply%20connected%3A%20for%20any%20loop%20%24%5Cgamma%24%2C%20contract%20it%20to%20the%20single%20connection%20point%20of%20the%20wedge%20sum.%20We%20can%20do%20it%20because%20%24%5Cgamma%24%20restricted%20on%20each%20sphere%20consists%20of%20loops%2C%20and%20each%20%24(n-1)%24-sphere%20is%20simply%20connected%20for%20%24n%5Cge%203%24.%20So%20our%20original%20space%20deformation%20retract%20to%20a%20simply%20connected%20space%2C%20thus%20it%20is%20simply%20connected.%0A%0A%20%20%20%20%23%23%20Problem%204%0A%0A%20%20%20%20**Let%20%24X%20%5Csubset%20%5Cmathbb%7BR%7D%5E3%24%20be%20the%20union%20of%20%24n%24%20lines%20through%20the%20origin.%20Compute%20%24%5Cpi_1(%5Cmathbb%7BR%7D%5E3%20-%20X)%24**.%0A%0A%20%20%20%20Solution%0A%0A%20%20%20%20This%20space%20is%20homotopic%20equivalent%20with%20%24S%5E2%24%20with%20%24n%24%20pairs%20of%20antipodal%20points%20removed%20through%20%24f%3A%20%5Cmathbb%7BR%7D%5E3%20%5Csetminus%20X%20%5Cto%20S%5E2%20%5Csetminus%20%5Clbrace%20n%20%5Ctext%7B%20pairs%20of%20points%7D%5Crbrace%24%2C%20sending%20%24x%20%5Cto%20x%2F%5C%7C%20x%20%5C%7C%24.%20Illustrate%20in%20lower%20dimension%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230226070640.png%22%20width%3D%22400%22%20%2F%3E%0A%0A%20%20%20%20(i.e.%20it's%20just%20the%20usual%20homotopy%20equivalence%20between%20%24%5Cmathbb%7BR%7D%5E3%20%5Csetminus%20%5Ctext%7Borigin%7D%20%5Cto%20S%5E2%24%20with%20some%20extra%20lines%20also%20removed)%0A%0A%20%20%20%20This%20new%20space%20is%20homotopy%20equivalent%20with%20%24D%5E2%20%5Csetminus%20%5Clbrace%202n-1%5Ctext%7B%20points%7D%5Crbrace%24%20('enlarge'%20one%20of%20the%20points)%2C%20which%20is%20homotopy%20equivalent%20with%20%24%5Cmathbb%7BR%7D%5E2%20%5Csetminus%20%5Clbrace%202n-1%20%5Ctext%7B%20points%7D%5Crbrace%24%2C%20whose%20fundamental%20group%20is%20the%20free%20group%20generated%20by%20%242n-1%24%20generators%20%24F_%7B2n-1%7D%24.%20Since%20the%20fundamental%20group%20is%20preserved%20by%20homotopy%20equivalence%2C%20%24%5Cpi_1(%5Cmathbb%7BR%7D%5E3%20%5Csetminus%20X)%20%5Ccong%20F_%7B2n-1%7D%24.%0A%0A%20%20%20%20%23%23%20Problem%208%0A%0A%20%20%20%20**Compute%20the%20fundamental%20group%20of%20the%20space%20obtained%20from%20two%20tori%20%24S%5E1%20%5Ctimes%20S%5E1%24%20by%20identifying%20a%20circle%20%24S%5E1%20%5Ctimes%20%5Clbrace%20x_0%20%5Crbrace%24%20in%20one%20torus%20with%20the%20corresponding%20circle%20%24S%5E1%20%5Ctimes%20%5Clbrace%20x_0%20%5Crbrace%24%20in%20the%20other%20torus**.%0A%0A%20%20%20%20Solution%0A%0A%20%20%20%20One%20embedding%20of%20this%20space%20is%20following%20(a%20big%20torus%20and%20a%20small%20torus%2C%20the%20small%20torus'%20outer%20longitude%20circle%20touches%20the%20inner%20longitude%20circle%20of%20the%20big%20torus)%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230226072615.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20This%20space%20is%20path%20connected%20so%20let%20us%20choose%20the%20base%20point%20to%20be%20on%20the%20identified%20circle%20(%24x%24%20on%20the%20diagram).%20Then%20the%20outer%20torus%20%24T_1%24%20has%20fundamental%20group%20%24%5Clangle%20a%2Cb%20%7C%20aba%5E%7B-1%7Db%5E%7B-1%7D%20%5Crangle%24%20(%24a%24%20is%20a%20longitude%2C%20%24b%24%20is%20a%20meridian)%2C%20the%20inner%20torus%20%24T_2%24%20has%20fundamental%20group%20%24%5Clangle%20a%2Cc%20%7C%20aca%5E%7B-1%7Dc%5E%7B-1%7D%20%5Crangle%24%20(the%20same%20%24a%24%20is%20also%20a%20longitude%20for%20%24T_2%24%20that%20passes%20%24x%24)%2C%20and%20their%20intersection%20%24T%24%20is%20a%20circle%20with%20fundamental%20group%20%24%5Clangle%20a%20%5Crangle%24%20(again%20it's%20the%20same%20%24a%24%20passing%20the%20same%20base%20point%20%24x%24).%0A%0A%20%20%20%20By%20Seifert-Van%20Kampen%20Theorem%20(call%20the%20entire%20space%20%24X%24)%2C%20we%20have%20%24%5Cpi_1(X%2Cx)%24%20%24%3D%20%5Clangle%20a%2Cb%2Cc%20%7C%20aba%5E%7B-1%7Db%5E%7B-1%7D%2C%20aca%5E%7B-1%7Dc%5E%7B-1%7D%2C%20R_S%5Crangle%24%20where%20%24R_S%24%20in%20our%20case%20is%20the%20relation%20(there%20is%20only%20one%20generator%20for%20%24T%24)%20%24%5Cvarphi_%7B1*%7D(a)%20%3D%20%5Cvarphi_%7B2*%7D(a)%24%20where%20%24%5Cvarphi_i%24's%20are%20the%20inclusion%20maps%20from%20%24T%20%5Cto%20T_i%24.%20In%20our%20case%20%24%5Cvarphi_%7B1*%7D(a)%24%20and%20%24%5Cvarphi_%7B2*%7D(a)%24%20are%20simply%20%24a%24%2C%20so%20this%20relation%20can%20be%20discard%2C%20and%20we%20get%20that%20%24%5Cpi_1(X)%20%5Ccong%20%5Cpi_1(X%2C%20x)%24%20%24%5Ccong%20%5Clangle%20a%2Cb%2Cc%20%7C%20aba%5E%7B-1%7Db%5E%7B-1%7D%2C%20aca%5E%7B-1%7Dc%5E%7B-1%7D%20%5Crangle%24.%0A%0A%20%20%20%20%23%23%20Problem%2014%0A%0A%20%20%20%20**Consider%20the%20quotient%20space%20of%20a%20cube%20%24I%5E3%24%20obtained%20by%20identifying%20each%20square%20face%20with%20the%20opposite%20square%20face%20via%20the%20right-handed%20screw%20motion%20consisting%20of%20a%20translation%20by%20one%20unit%20in%20the%20direction%20perpendicular%20to%20the%20face%20combined%20with%20a%20one-quarter%20twist%20of%20the%20face%20about%20its%20center%20point.%20Show%20this%20quotient%20space%20%24X%24%20is%20a%20cell%20complex%20with%20two%20%240%24-cells%2C%20four%20%241%24-cells%2C%20three%20%242%24-cells%2C%20and%20one%20%243%24-cell.%20Using%20this%20structure%2C%20show%20that%20%24%5Cpi_1(X)%24%20is%20the%20quaternion%20group%20%24%5Clbrace%20%5Cpm%201%2C%20%5Cpm%20i%2C%20%5Cpm%20j%20%2C%20%5Cpm%20k%20%5Crbrace%24%2C%20of%20order%20eight**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Start%20with%20a%20cube%20%24I%5E3%24%20and%20do%20the%20identification%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230226155418.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20We%20apply%20the%20procedure%20on%20the%20book%20(p.%2050~51)%20to%20calculate%20the%20fundamental%20group.%0A%0A%20%20%20%20We%20start%20with%20the%20%241%24-skeleton%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230226162554.png%22%20width%3D%22400%22%20%2F%3E%0A%0A%20%20%20%20So%20%24X%5E1%24%20can%20be%20seen%20as%20wedge%20sum%20of%20%243%24%20circles%2C%20and%20has%20fundamental%20group%20%24%5Cpi_1(X%5E1%2Cu)%20%5Ccong%20%5Clangle%20cb%2C%20cd%2C%20ca%5E%7B-1%7D%5Crangle%24%20free%20group%20generated%20by%20%243%24%20generic%20loops%20in%20the%20space%20based%20at%20%24u%24%20(the%20choice%20of%20base%20point%20and%20generators%20are%20of%20course%20not%20unique%2C%20though).%0A%0A%20%20%20%20Now%20we%20need%20to%20attach%20the%20three%20%242%24-cells%20to%20it.%20By%20Proposition%201.26%2C%20we%20get%20that%20%24%5Cpi_1(X%5E2%2C%20u)%24%20%24%5Ccong%20%5Clangle%20cb%2Ccd%2Cca%5E%7B-1%7D%20%7C%20abcd%2C%20ca%5E%7B-1%7Db%5E%7B-1%7Dd%2C%20d%5E%7B-1%7Da%5E%7B-1%7Dcb%20%5Crangle%24%2C%20using%20the%20attaching%20map%20illustrated%20in%20the%20right-most%20diagram%20in%20the%20first%20picture%20(i.e.%20%24A%24%20is%20attached%20through%20%24abcd%24%2C%20etc.%2C%20choose%20the%20loops%20to%20be%20based%20at%20%24u%24%20so%20that%20we%20do%20not%20need%20to%20worry%20about%20change%20of%20base%20points).%20Again%20by%20that%20proposition%2C%20this%20is%20(isomorphic%20to)%20the%20fundamental%20group%20of%20%24X%24.%0A%0A%20%20%20%20Now%20we%20need%20to%20show%20this%20presentation%20gives%20us%20the%20quaternion%20group.%20Rewrite%20the%20presentation%20as%20follow%3A%20%24%24%5Clangle%20cb%2Ccd%2Cca%5E%7B-1%7D%20%7C%20ac%5E%7B-1%7Dcbcd%2C%20ca%5E%7B-1%7Db%5E%7B-1%7Dc%5E%7B-1%7Dcd%2C%20d%5E%7B-1%7Dc%5E%7B-1%7Dca%5E%7B-1%7Dcb%20%5Crangle%24%24%20by%20group%20operation%20and%20inserting%20%24c%5E%7B-1%7Dc%24%20to%20the%20relations.%20If%20we%20write%20%24cb%20%3D%20i%2C%20cd%20%3D%20j%2C%20ca%5E%7B-1%7D%20%3D%20k%24%2C%20we%20get%20that%20%24%24%5Clangle%20i%2Cj%2Ck%20%7C%20k%5E%7B-1%7Dij%2C%20ki%5E%7B-1%7Dj%2Cj%5E%7B-1%7Dki%5Crangle%24%24%20or%20%24%24%5Clangle%20i%2Cj%2Ck%20%7C%20i%5E2%20%3D%20j%5E2%20%3D%20k%5E2%20%3D%20ijk%5Crangle%24%24%20which%20is%20a%20presentation%20of%20the%20quaternion%20group.%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%20Spring%20Semester%20Homework%205%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%23%20Problem%204%0A%0A%20%20%20%20**Construct%20a%20simply-connected%20covering%20space%20of%20the%20space%20%24X%20%5Csubset%20%5Cmathbb%7BR%7D%5E3%24%20that%20is%20the%20union%20of%20a%20sphere%20and%20a%20diameter.%20Do%20the%20same%20when%20%24X%24%20is%20the%20union%20of%20a%20sphere%20and%20a%20circle%20intersecting%20it%20in%20two%20points**.%0A%0A%20%20%20%20Solution%0A%0A%20%20%20%20Consider%20%24%5Ctilde%7BX%7D%24%20to%20be%20the%20following%20space%20(in%20%24%5Cmathbb%7BR%7D%5E3%24)%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230309123311.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20Which%20is%20an%20infinite%20copy%20of%20a%20sphere%20and%20a%20line%20segment.%20This%20space%20is%20homotopic%20equivalent%20with%20an%20infinite%20wedge%20sum%20of%20%24S%5E2%24%20which%20is%20simply-connected.%20Consider%20%24p%3A%20%5Ctilde%7BX%7D%20%5Cto%20X%24%20to%20be%20the%20map%20that%20maps%20points%20on%20spheres%20to%20points%20on%20the%20sphere%3B%20points%20on%20line%20segments%20to%20points%20on%20the%20diameter%3B%20and%20points%20on%20the%20intersection%20of%20sphere%20and%20line%20segments%20to%20points%20on%20the%20intersection%20of%20sphere%20and%20diameter.%20Illustrated%20as%20below%2C%20image%2Fpre-images%20are%20corresponded%20in%20the%20same%20color%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230309124205.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20Locally%20this%20%24p%24%20works%20like%20a%20projection%20(the%20line%20segment%20is%20not%20inside%20the%20sphere%20through)%2C%20so%20it%20is%20an%20%24%5Cinfty%24-to-%20%241%24%20covering%20map.%0A%0A%20%20%20%20For%20the%20second%20part%2C%20consider%20the%20following%20space%20as%20%24%5Ctilde%7BX%7D%24%20and%20correspondence%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230309130016.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20(The%20graph%20in%20fact%20looks%20like%20the%20Cayley%20graph%20as%20shown%20on%20p.%2077.%20The%20places%20like%20the%20indicated%20one%20above%20are%20not%20connecting%2C%20so%20it%20will%20not%20create%20a%20non-trivial%20loop)%20Again%20this%20%24%5Ctilde%7BX%7D%24%20is%20simply-connected.%20%24p%24%20maps%20%24%5Ctilde%7BX%7D%24%20onto%20%24X%24%20in%20a%20similar%20manner%20as%20the%20above%20%24p%24%2C%20it%20is%20also%20an%20%24%5Cinfty%24-to-%20%241%24%20covering%20map.%0A%0A%20%20%20%20%23%23%20Problem%208%0A%0A%20%20%20%20**Let%20%24%5Ctilde%7BX%7D%24%20and%20%24%5Ctilde%7BY%7D%24%20be%20simply-connected%20covering%20spaces%20of%20the%20path-connected%2C%20locally%20path-connected%20spaces%20%24X%24%20and%20%24Y%24.%20Show%20that%20if%20%24X%20%5Csimeq%20Y%24%20then%20%24%5Ctilde%7BX%7D%20%5Csimeq%20%5Ctilde%7BY%7D%24**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Let%20%24f%3A%20X%20%5Cto%20Y%24%20and%20%24g%3A%20Y%20%5Cto%20X%24%20be%20the%20maps%20that%20realize%20the%20homotopy%20equivalence%20between%20%24X%24%20and%20%24Y%24.%20Denote%20%24p%3A%20%5Ctilde%7BX%7D%20%5Cto%20X%24%20and%20%24q%3A%20%5Ctilde%7BY%7D%20%5Cto%20Y%24%20to%20be%20the%20covering%20maps.%20Consider%20%24f%20%5Ccirc%20p%3A%20%5Ctilde%7BX%7D%20%5Cto%20X%20%5Cto%20Y%24.%0A%0A%20%20%20%20By%20Proposition%201.33%20(we%20can%20apply%20it%20because%20%24X%24%20is%20locally%20path-connected%20implies%20%24%5Ctilde%7BX%7D%24%20is%20locally%20path%20connected%2C%20and%20%24%5Ctilde%7BX%7D%24%20is%20simply-connected%20implies%20it%20is%20path-connected)%2C%20a%20lift%20%24%5Cwidetilde%7Bf%20%5Ccirc%20p%7D%3A%20%5Ctilde%7BX%7D%20%5Cto%20%5Ctilde%7BY%7D%24%20exists%20if%20and%20only%20if%20%24(f%20%5Ccirc%20p)_*(%5Cpi_1(%5Ctilde%7BX%7D))%20%5Csubset%20q_*(%5Cpi_1(%5Ctilde%7BY%7D))%24-%20the%20latter%20is%20true%20because%20of%20the%20simply-connected%20assumption%20so%20these%20are%20both%20trivial%20groups.%20For%20simplicity%20let%20us%20denote%20%24%5Cwidetilde%7Bf%20%5Ccirc%20p%7D%24%20as%20%24%5Ctilde%7Bf%7D%24%20(and%20by%20lifting%20we%20mean%20that%20%24q%20%5Ccirc%20%5Ctilde%7Bf%7D%20%3D%20f%5Ccirc%20p%24).%20Similarly%2C%20use%20%24g%20%5Ccirc%20q%24%20and%20%24p%24%2C%20we%20can%20get%20a%20lift%20%24%5Ctilde%7Bg%7D%3A%20%5Ctilde%7BY%7D%20%5Cto%20%20%5Ctilde%7BX%7D%24%20(and%20that%20%24p%20%5Ccirc%20%5Ctilde%7Bg%7D%20%3D%20g%20%5Ccirc%20q%24).%0A%0A%20%20%20%20We%20need%20to%20show%20that%20%24%5Ctilde%7Bf%7D%24%20is%20a%20homotopy%20equivalence%2C%20we%20show%20it%20by%20showing%20that%20we%20can%20find%20some%20%24%5Cvarphi%24%20and%20%24%5Cpsi%24%20so%20that%20%24%5Cvarphi%20%5Ccirc%20%5Ctilde%7Bf%7D%24%20and%20%24%5Ctilde%7Bf%7D%20%5Ccirc%20%5Cpsi%24%20are%20both%20homotopic%20to%20identity%20(and%20the%20result%20follows%20from%200.11%20as%20hinted).%0A%0A%20%20%20%20Notice%20by%20above%2C%20we%20have%20that%20%24q%5Ccirc%20%5Ctilde%7Bf%7D%20%5Ccirc%5Ctilde%7Bg%7D%24%20%24%3D%20f%5Ccirc%20p%20%5Ccirc%20%5Ctilde%7Bg%7D%24%20%24%3D%20f%5Ccirc%20g%20%5Ccirc%20q%24%2C%20which%20means%20that%20%24%5Ctilde%7Bf%7D%20%5Ccirc%20%5Ctilde%7Bg%7D%24%20is%20a%20lift%20of%20%24f%20%5Ccirc%20g%20%5Ccirc%20q%24.%20We%20also%20know%20from%20assumption%20that%20%24f%20%5Ccirc%20g%20%5Csimeq%20%5Ctext%7Bid%7D%24%20thus%20%24f%20%5Ccirc%20g%20%5Ccirc%20q%20%5Csimeq%20q%24%20(proved%20in%20Homework%2001%20in%20Fall%20semester).%0A%0A%20%20%20%20Let%20%24H%24%20be%20the%20homotopy%20that%20realizes%20%24f%20%5Ccirc%20g%20%5Ccirc%20q%20%5Csimeq%20q%24%20(say%20with%20%24h_0%20%3D%20f%20%5Ccirc%20g%20%5Ccirc%20q%24%20and%20%24h_1%20%3D%20q%24)%2C%20then%20by%20Homotopy%20Lifting%20Property%2C%20we%20have%20a%20unique%20%24%5Ctilde%7BH%7D%24%20with%20%24%5Ctilde%7Bh%7D_0%20%3D%20%5Ctilde%7Bf%7D%20%5Ccirc%20%5Ctilde%7Bg%7D%24.%20Consider%20%24%5Ctilde%7Bh%7D_1%24%2C%20it%20is%20some%20lift%20of%20%24q%24.%20Since%20the%20corresponding%20covering%20map%20is%20%24q%24%2C%20by%20lifting%20we%20mean%20that%20%24q%5Ccirc%20%5Ctilde%7Bh%7D_1%20%3D%20q%24.%20Thus%20for%20any%20%24%5Ctilde%7By%7D%20%5Cin%20%5Ctilde%7BY%7D%24%2C%20we%20have%20%24q(%5Ctilde%7Bh%7D_1(%5Ctilde%7By%7D))%20%3D%20q(%5Ctilde%7By%7D)%24%2C%20which%20means%20%24%5Ctilde%7Bh%7D_1(%5Ctilde%7By%7D)%24%20and%20%24%5Ctilde%7By%7D%24%20are%20in%20some%20same%20fiber.%20Since%20%24%5Ctilde%7BY%7D%24%20is%20simply-connected%2C%20%24q%24%20is%20a%20normal%20covering%2C%20thus%20there%20exists%20some%20deck%20transformation%20%24u%3A%20%5Ctilde%7BY%7D%20%5Cto%20%5Ctilde%7BY%7D%24%20sending%20%24%5Ctilde%7Bh%7D_1(%5Ctilde%7By%7D)%24%20to%20%24%5Ctilde%7By%7D%24.%20In%20particular%20%24%5Ctilde%7Bh%7D_1%5Ccirc%20u%24%20maps%20%24%5Ctilde%7Bh%7D_1(%5Ctilde%7By%7D)%24%20to%20itself.%0A%0A%20%20%20%20Notice%20now%20%24%5Ctilde%7Bh%7D_1%5Ccirc%20u%24%20is%20a%20lift%20of%20%24q%24%20by%20definition%20of%20a%20deck%20transformation%20(%24q%5Ccirc%20%5Ctilde%7Bh%7D_1%20%5Ccirc%20u%24%20%24%3Dq%5Ccirc%20u%24%20%24%3D%20q%24)%3B%20%24%5Ctext%7Bid%7D%24%20(on%20%24%5Ctilde%7BY%7D%24)%20is%20of%20course%20also%20a%20lift%20of%20%24q%24%20and%20they%20agree%20at%20%24%5Ctilde%7Bh%7D_1(%5Ctilde%7By%7D)%24%2C%20thus%20Unique%20Lifting%20Property%20applies%2C%20we%20get%20that%20%24%5Ctilde%7Bh%7D_1%20%5Ccirc%20u%20%3D%20%5Ctext%7Bid%7D%24.%20Finally%2C%20define%20%24%5Cpsi%24%20to%20be%20%24%5Ctilde%7Bg%7D%5Ccirc%20u%24%2C%20we%20have%20found%20%24%5Cpsi%24%20so%20that%20%24%5Ctilde%7Bf%7D%20%5Ccirc%20%5Cpsi%24%20%24%3D%5Ctilde%7Bf%7D%5Ccirc%5Ctilde%7Bg%7D%20%5Ccirc%20u%24%20%24%5Csimeq%20%5Ctilde%7Bh%7D_1%20%5Ccirc%20u%24%20%24%3D%20%5Ctext%7Bid%7D%24.%0A%0A%20%20%20%20Similarly%2C%20use%20that%20%24p%5Ccirc%20%5Ctilde%7Bg%7D%20%5Ccirc%20%5Ctilde%7Bf%7D%24%20%24%3D%20g%5Ccirc%20q%20%5Ccirc%20%5Ctilde%7Bf%7D%24%20%24%3D%20g%20%5Ccirc%20f%20%5Ccirc%20p%24%20with%20a%20similar%20argument%20(but%20possibly%20different%20order%20of%20composing%20functions%20in%20the%20steps)%2C%20we%20can%20find%20the%20desired%20%24%5Cvarphi%24%20that%20realizes%20%24%5Cvarphi%20%5Ccirc%20f%20%5Csimeq%20%5Ctext%7Bid%7D%24.%20Combine%20everything%20we%20have%2C%20we%20get%20%24%5Ctilde%7BX%7D%20%5Csimeq%20%5Ctilde%7BY%7D%24.%0A%0A%20%20%20%20%23%23%20Problem%209%0A%0A%20%20%20%20**Show%20that%20if%20a%20path-connected%2C%20locally%20path-connected%20space%20%24X%24%20has%20%24%5Cpi_1(X)%24%20finite%2C%20then%20every%20map%20%24X%20%5Cto%20S%5E1%24%20is%20null-homotopic**.%20*Hint%3A%20Use%20the%20covering%20space%20%24%5Cmathbb%7BR%7D%20%5Cto%20S%5E1%24*.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Let%20%24f%3A%20X%5Cto%20S%5E1%24%20be%20an%20arbitrary%20(continuous)%20map.%20Then%20it%20induces%20homomorphism%20%24f_*%3A%20%5Cpi_1(X)%20%5Cto%20%5Cpi_1(S%5E1)%24%20which%20by%20assumption%20is%20a%20homomorphism%20from%20a%20finite%20group%20to%20%24%5Cmathbb%7BZ%7D%24%2C%20which%20must%20be%20trivial%20(%24%5Cmathbb%7BZ%7D%24%20does%20not%20have%20a%20non-trivial%20finite%20subgroup).%0A%0A%20%20%20%20Apply%20Proposition%201.33%2C%20using%20%24p%3A%20%5Cmathbb%7BR%7D%20%5Cto%20S%5E1%24%20the%20standard%20universal%20covering%20and%20the%20assumption%20that%20%24X%24%20is%20path-connected%20and%20locally%20path-connected.%20The%20inclusion%20condition%20is%20automatically%20true%20because%20%24f_*%24%20is%20trivial%20as%20argued.%20Thus%20we%20can%20always%20get%20a%20lift%20%24%5Ctilde%7Bf%7D%3A%20X%20%5Cto%20%5Cmathbb%7BR%7D%24%2C%20with%20%24p%20%5Ccirc%20%5Ctilde%7Bf%7D%20%3D%20f%24.%0A%0A%20%20%20%20Since%20%24%5Cmathbb%7BR%7D%24%20is%20contractible%2C%20%24%5Ctilde%7Bf%7D%24%20is%20null-homotopic%2C%20thus%20any%20composition%20of%20it%20in%20the%20form%20of%20%24g%20%5Ccirc%20%5Ctilde%7Bf%7D%24%20is%20null-homotopic%2C%20which%20in%20particular%20means%20%24f%24%20is%20null-homotopic%2C%20as%20desired.%0A%0A%20%20%20%20%23%23%20Problem%2010%0A%0A%20%20%20%20**Find%20all%20the%20connected%20%242%24-sheeted%20and%20%243%24-sheeted%20covering%20spaces%20of%20%24S%5E1%20%5Cvee%20S%5E1%24%2C%20up%20to%20isomorphism%20of%20covering%20spaces%20without%20base-points**.%0A%0A%20%20%20%20Solution%0A%0A%20%20%20%20View%20%24S%5E1%20%5Cvee%20S%5E1%24%20as%20a%20graph%20with%20one%20vertex%20and%20two%20edges%20%24a%2C%20b%24%20(oriented).%20This%20vertex%20has%20both%20edges%20%24a%24%20and%20%24b%24%20coming%20in%20and%20coming%20out.%20So%20for%20the%20%242%24-sheeted%20cover%20we%20should%20have%20two%20vertices%2C%20two%20edge%20%24a%24%20and%20two%20edge%20%24b%24%2C%20with%20each%20vertex%20having%20both%20%24a%2Cb%24%20coming%20in%20and%20coming%20out.%20We%20can%20also%20assume%2Fdraw%20the%20graph%20without%20edges%20go%20over%20each%20other%20because%20we%20are%20classifying%20up%20to%20isomorphism%2Fhomeomorphism.%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230312085113.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20Iterate%20all%20the%20possibility.%20Start%20with%20one%20edge%20%24a%24%20coming%20out%20of%20one%20of%20the%20vertex%2C%20then%20it%20either%3A%0A%0A%20%20%20%201.%20Gets%20back%20to%20the%20same%20vertex.%20Then%20the%20other%20vertex%20must%20also%20have%20an%20edge%20%24a%24%20coming%20out%20and%20coming%20in%20back%20(otherwise%20the%20first%20vertex%20will%20have%20two%20%24a%24%20coming%20in).%20Then%20consider%20one%20edge%20%24b%24%20coming%20out%20of%20the%20first%20vertex%3A%0A%20%20%20%20%091.%20If%20it%20comes%20back%20to%20the%20same%20vertex%2C%20then%20the%20other%20vertex%20also%20have%20(another)%20edge%20%24b%24%20coming%20out%20and%20comes%20back%20in%2C%20makes%20the%20graph%20not%20connected%20(graph%200)%3B%0A%20%20%20%20%092.%20If%20it%20goes%20to%20the%20other%20vertex%2C%20then%20the%20other%20edge%20%24b%24%20must%20come%20out%20of%20the%20second%20vertex%20and%20come%20in%20to%20the%20first%20vertex%2C%20we%20get%20graph%203%3B%0A%20%20%20%202.%20Otherwise%20it%20goes%20in%20to%20the%20other%20vertex%2C%20use%20the%20same%20reasoning%20above%3A%0A%20%20%20%20%091.%20If%20edge%20%24b%24%20comes%20out%20of%2Fin%20to%20different%20vertices%2C%20we%20get%20graph%201%3B%0A%20%20%20%20%092.%20Otherwise%20we%20get%20graph%202.%0A%0A%20%20%20%20%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230312085538.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20%20%20%20%20(The%20labeled%20orientation%20and%20the%20position%20ordering%20of%20the%20edges%20are%20not%20important%20in%20%242%24-sheeted%20case)%20These%20complete%20the%20classification%2C%20there%20are%20three%20possibilities.%0A%0A%20%20%20%20For%20%243%24-sheeted%20covering%2C%20we%20can%20follow%20the%20same%20reasoning%20-%20the%20only%20difference%20is%20now%20we%20have%20three%20vertices%20and%20six%20edges%20(three%20%24a%24%20and%20three%20%24b%24).%20Start%20with%20all%20edges%20come%20out%20of%2Fin%20to%20different%20vertices%2C%20and%20iterate%20all%20possibilities.%20We'll%20omit%20the%20word%20description%20here%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230312111230.png%22%20width%3D%22600%22%20%2F%3E%0A%0A%20%20%20%20Notice%20that%20unlike%20the%20%242%24-sheeted%20cover%2C%20orientations%20do%20matter%20in%20some%20cases%2C%20so%20for%20example%2C%20the%20first%20two%20graphs%20are%20not%20isomorphic%20(we%20can%20change%20orientations%20all%20together%2C%20but%20not%20changing%20orientation%20of%20%24b%24's%20without%20changing%20%24a%24's).%20In%20total%20there%20are%20seven%20possibilities.%0A%0A%20%20%20%20%23%23%20Problem%203%0A%0A%20%20%20%20**Prove%20the%20Borsuk-Ulam%20Theorem%20by%20the%20following%20argument.%20Suppose%20on%20the%20contrary%20that%20%24f%3A%20S%5En%20%5Cto%20%5Cmathbb%7BR%7D%5En%24%20satisfies%20%24f(x)%20%5Cne%20f(-x)%24%20for%20all%20%24x%24.%20Then%20define%20%24g%3A%20S%5En%20%5Cto%20S%5E%7Bn-1%7D%24%20by%20%24g(x)%20%3D%20(f(x)%20-%20f(-x))%2F%7Cf(x)%20-%20f(-x)%7C%24%2C%20so%20%24g(-x)%20%3D%20-g(x)%24%20and%20%24g%24%20induces%20a%20map%20%24%5Cmathbb%7BR%7DP%5En%20%5Cto%20%5Cmathbb%7BR%7DP%5E%7Bn-1%7D%24.%20Show%20that%20part%20(a)%20applies%20to%20this%20map**.%0A%0A%20%20%20%20%3E%20Part%20(a)%0A%20%20%20%20%3E%0A%20%20%20%20%3E%20There%20is%20no%20map%20%24%5Cmathbb%7BR%7DP%5En%20%5Cto%20%5Cmathbb%7BR%7DP%5Em%24%20inducing%20a%20non-trivial%20map%20%24H%5E1(%5Cmathbb%7BR%7DP%5Em%3B%5Cmathbb%7BZ%7D_2)%20%5Cto%20H%5E1(%5Cmathbb%7BR%7DP%5En%3B%5Cmathbb%7BZ%7D_2)%24%20if%20%24n%20%3E%20m%24.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Let%20us%20call%20the%20induced%20map%20%24h%3A%20%5Cmathbb%7BR%7DP%5En%20%5Cto%20%5Cmathbb%7BR%7DP%5E%7Bn-1%7D%24.%20It%20would%20induce%20homomorphism%20%24h_*%3A%20%5Cpi_1(%5Cmathbb%7BR%7DP%5En)%20%5Cto%20%5Cpi_1(%5Cmathbb%7BR%7DP%5E%7Bn-1%7D)%24.%20We%20would%20show%20that%20%24h_*%24%20is%20non-trivial.%20If%20that%20is%20true%2C%20then%20since%20these%20fundamental%20groups%20are%20both%20abelian%20thus%20%24h_*%24%20also%20gives%20us%20non-trivial%20homomorphism%20between%20the%20first%20homology%20groups%2C%20which%20will%20give%20us%20non-trivial%20dual%20between%20first%20cohomology%20groups%20(with%20%24%5Cmathbb%7BZ%7D_2%24%20coefficient)%2C%20which%20contradict%20part%20(a).%0A%0A%20%20%20%20Pick%20%24%5Cvarphi%24%20so%20that%20it%20is%20a%20non-trivial%20loop%20%24I%5Cto%20%5Cmathbb%7BR%7DP%5En%24.%20Consider%20the%20%242%24-sheeted%20covering%20%24S%5En%20%5Cto%20%5Cmathbb%7BR%7DP%5En%24%2C%20we%20can%20lift%20%24%5Cvarphi%24%20to%20%24%5Ctilde%7B%5Cvarphi%7D%3A%20I%20%5Cto%20S%5En%24.%20The%20endpoints%20%24%5Ctilde%7B%5Cvarphi%7D(0)%24%20and%20%24%5Ctilde%7B%5Cvarphi%7D(1)%24%20must%20be%20antipodal%20(else%20the%20endpoints%20would%20be%20same%20and%20%24%5Cvarphi%24%20is%20trivial)%2C%20in%20particular%20we%20have%20%24g%5Ccirc%20%5Ctilde%7B%5Cvarphi%7D%3AI%20%5Cto%20S%5En%20%5Cto%20S%5E%7Bn-1%7D%24%20also%20has%20endpoints%20being%20antipodal.%20Consider%20the%20following%20diagram%3A%20%24%24%5Cbegin%7Barray%7D%7Bccccc%7D%26%26S%5En%26%5Cxrightarrow%7Bg%7D%26S%5E%7Bn-1%7D%5C%5C%26%5Cnearrow_%7B%5Ctilde%7B%5Cvarphi%7D%7D%26%5Cdownarrow%5Ep%26%26%5Cdownarrow%5Ep%5C%5CI%26%5Cxrightarrow%7B%5Cvarphi%7D%26%5Cmathbb%7BR%7DP%5En%26%5Cxrightarrow%7Bh%7D%26%5Cmathbb%7BR%7DP%5E%7Bn-1%7D%5Cend%7Barray%7D%24%24%20It%20is%20commutative%20because%20of%20lifting%20and%20the%20construction%20of%20%24g%24.%20Thus%20%24g%5Ccirc%20%5Ctilde%7B%5Cvarphi%7D%24%20is%20a%20lift%20of%20%24h%20%5Ccirc%20%5Cvarphi%24.%20Since%20%24g%5Ccirc%20%5Ctilde%7B%5Cvarphi%7D%24%20is%20not%20a%20loop%20as%20discussed%2C%20%24%5Bh%5Ccirc%20%5Cvarphi%5D%20%3D%20h_*(%5B%5Cvarphi%5D)%24%20is%20not%20trivial%2C%20which%20in%20particular%20means%20%24h_*%24%20is%20not%20trivial.%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%20Spring%20Semester%20Homework%206%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%23%20Problem%2012%0A%0A%20%20%20%20**Show%20that%20an%20%24n%24-connected%2C%20%24n%24-dimensional%20CW-complex%20is%20contractible**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Call%20this%20space%20%24X%24.%20Consider%20the%20Compression%20Lemma%20applied%20with%20%24Y%20%3D%20X%24%2C%20%24B%20%3D%20x_0%24%20and%20%24A%24%20be%20a%20sub-complex%20that%20contains%20%24x_0%24.%20The%20assumption%20%22%E2%80%A6%20%24%5Cpi_n(Y%2C%20B%2C%20y_0)%20%3D%200%24%20for%20all%20%24y_0%20%5Cin%20B%24%20%22%20is%20true%20because%20of%20the%20%24n%24-connectedness%20assumption.%20Thus%20the%20lemma%20applies%20and%20in%20particular%20we%20get%20that%20the%20identity%20map%20%24%5Ctext%7Bid%7D%3A%20(X%2C%20A)%20%3D%20X%20%5Cto%20(X%2C%20x_0)%24%20is%20homotopic%20(relative%20to%20%24A%24)%20to%20a%20map%20%24X%20%5Cto%20x_0%24%20(a%20constant%20map)%2C%20which%20by%20definition%20means%20%24X%24%20is%20contractible%20(identity%20is%20null-homotopic).%0A%0A%20%20%20%20%23%23%20Problem%2014%0A%0A%20%20%20%20**Use%20cellular%20approximation%20to%20show%20that%20the%20%24n%24-skeletons%20of%20homotopy%20equivalent%20CW-complexes%20without%20cells%20of%20dimension%20%24n%2B1%24%20are%20also%20homotopy%20equivalent**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Suppose%20%24X%2C%20Y%24%20are%20two%20homotopy%20equivalent%20CW-complexes%20without%20cells%20of%20dimension%20%24n%2B1%24.%20Let%20%24f%3A%20X%20%5Cto%20Y%24%20and%20%24g%3A%20Y%5Cto%20X%24%20be%20the%20two%20functions%20that%20realize%20the%20homotopy%20equivalence.%20By%20cellular%20approximation%2C%20we%20may%20assume%20that%20%24f%24%20and%20%24g%24%20are%20cellular%20maps%20(compositions%20of%20homotopic%20maps%20are%20still%20homotopic).%20Claim%3A%20%24f%24%20and%20%24g%24%20restricted%20to%20the%20%24n%24-skeletons%20of%20%24X%24%20and%20%24Y%24%20are%20the%20functions%20that%20realize%20the%20homotopy%20equivalence%20of%20the%20%24n%24-skeletons%3A%0A%0A%20%20%20%20Let%20%24H%3A%20Y%20%5Ctimes%20I%20%5Cto%20Y%24%20be%20the%20homotopy%20between%20%24f%20%5Ccirc%20g%24%20and%20%24%5Ctext%7Bid%7D%24.%20By%20viewing%20%24I%24%20as%20a%20one-cell%2C%20there%20is%20a%20natural%20CW%20structure%20on%20%24Y%20%5Ctimes%20I%24.%20Since%20the%20restriction%20of%20%24H%24%20on%20%24Y%20%5Ctimes%20%5Clbrace%200%2C%201%20%5Crbrace%24%20(which%20is%20%24f%5Ccirc%20g%24%20and%20%24%5Ctext%7Bid%7D%24)%20is%20a%20cellular%20map%2C%20by%20cellular%20approximation%2C%20we%20can%20get%20a%20cellular%20map%20%24H'%3A%20Y%20%5Ctimes%20I%20%5Cto%20Y%24%20that%20is%20homotopic%20to%20%24H%24%20relative%20to%20%24Y%20%5Ctimes%20%5Clbrace%200%20%2C%201%20%5Crbrace%24%2C%20i.e.%20we%20get%20a%20homotopy%20(that%20itself%20is%20a%20cellular%20map)%20between%20%24f%20%5Ccirc%20g%24%20and%20%24%5Ctext%7Bid%7D%24.%20Now%20restrict%20%24H'%24%20to%20%24Y%5E%7B(n)%7D%5Ctimes%20I%24%2C%20we%20have%20%24H'(Y%5E%7B(n)%7D%20%5Ctimes%20I)%20%5Csubseteq%20H'((Y%20%5Ctimes%20I)%5E%7B(n%2B1)%7D)%24%20%24%5Csubseteq%20Y%5E%7B(n%2B1)%7D%24%20%24%3D%20Y%5E%7B(n)%7D%24%20(first%20step%20is%20due%20to%20product%20of%20CW-complexes%2C%20second%20step%20is%20due%20to%20cellular%20map%2C%20third%20step%20is%20due%20to%20there%20is%20no%20%24n%2B1%24-cell%20in%20%24Y%24).%0A%0A%20%20%20%20So%20this%20restriction%20is%20a%20(continuous)%20map%20between%20%24Y%5E%7B(n)%7D%20%5Ctimes%20I%24%20and%20%24Y%5E%7B(n)%7D%24%3B%20its%20restriction%20to%20%24t%20%3D%200%24%20is%20%24f%20%5Ccirc%20g%24%20restricted%20to%20%24Y%5E%7B(n)%7D%24%20which%20is%20the%20same%20as%20%24f%7C_%7BX%5E%7B(n)%7D%7D%20%5Ccirc%20g%7C_%7BY%5E%7B(n)%7D%7D%24%20(by%20cellular%20map%20property)%2C%20and%20its%20restriction%20to%20%24t%20%3D%201%24%20is%20identity%20restricted%20to%20%24Y%5E%7B(n)%7D%24.%20So%20this%20restriction%20realizes%20that%20%24f%7C_%7BX%5E%7B(n)%7D%7D%20%5Ccirc%20g%7C_%7BY%5E%7B(n)%7D%7D%24%20and%20%24%5Ctext%7Bid%7D%24%20are%20homotopic.%20Similarly%2C%20we%20can%20get%20the%20homotopy%20between%20%24g%7C_%7BY%5E%7B(n)%7D%7D%20%5Ccirc%20f%7C_%7BX%5E%7B(n)%7D%7D%24%20and%20%24%5Ctext%7Bid%7D%24%2C%20so%20we%20proved%20the%20claim.%0A%0A%20%20%20%20%23%23%20Problem%0A%0A%20%20%20%20**Show%20that%20the%20suspension%20homomorphism%20%24%5Cpi_n(X)%20%5Cto%20%5Cpi_%7Bn%2B1%7D(%5CSigma%20X)%24%20defined%20by%20%24%5Bf%5D%20%5Cmapsto%20%5B%5CSigma%20f%5D%24%20is%20well-defined**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20First%20we%20show%20the%20map%20is%20well-defined.%20We%20need%20to%20show%20that%20given%20two%20maps%20%24f%2Cg%3A%20S%5En%20%5Cto%20X%24%20that%20are%20homotopic%2C%20their%20suspensions%20%24%5CSigma%20f%2C%20%5CSigma%20g%3A%20S%5E%7Bn%2B1%7D%20%5Cto%20%5CSigma%20X%24%20are%20also%20homotopic.%20This%20is%20straight%20forward%3A%20given%20%24H%24%20a%20homotopy%20between%20%24f%24%20and%20%24g%24%2C%20consider%20%24H%20%5Ctimes%20%5Ctext%7Bid%7D_I%3A%20S%5En%20%5Ctimes%20I%20%5Ctimes%20I%20%5Cto%20X%20%5Ctimes%20I%24%20and%20then%20the%20quotient%20%24%5CSigma%20H%3A%20S%5En%5Ctimes%20I%20%5Ctimes%20I%2F%5Csim%20%5Cto%20X%5Ctimes%20I%2F%5Csim%24%20which%20is%20the%20same%20as%20%24%5CSigma%20H%3A%20S%5E%7Bn%2B1%7D%5Ctimes%20I%20%5Cto%20%5CSigma%20X%24%3B%20it%20is%20continuous%20and%20by%20construction%20has%20%24%5CSigma%20f%24%20and%20%24%5CSigma%20g%24%20as%20it's%20restriction%20to%20the%20end-points%2C%20so%20%24%5CSigma%20f%24%20and%20%24%5CSigma%20g%24%20are%20homotopic.%0A%0A%20%20%20%20To%20show%20the%20map%20(denote%20it%20by%20%24%5CSigma%24)%20is%20a%20homomorphism%2C%20we%20need%20that%20for%20%24%5Bf%5D%2C%20%5Bg%5D%24%20in%20%24%5Cpi_n(X)%24%20(so%20that%20%24%5Bf%5D%5Ccdot%20%5Bg%5D%20%3D%20%5Bf%20%5Ccdot%20g%5D%24)%2C%20we%20have%20%24%5CSigma(%5Bf%5D)%5Ccdot%20%5CSigma(%5Bg%5D)%20%3D%20%5CSigma(%5Bf%5Ccdot%20g%5D)%24%2C%20or%20%24%5B%5CSigma%20f%5D%5Ccdot%20%5B%5CSigma%20g%5D%20%3D%20%5B%5CSigma(f%20%5Ccdot%20g)%5D%24.%20We%20will%20show%20in%20fact%20that%20%24%5CSigma%20f%20%5Ccdot%20%5CSigma%20g%20%3D%20%5CSigma(f%20%5Ccdot%20g)%24.%20It's%20rather%20straight%20forward%20as%20illustrated%3A%0A%0A%20%20%20%20%3Cimg%20src%3D%22public%2FPasted%20image%2020230413114808.png%22%20width%3D%22600%22%20%2F%3E%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%20Spring%20Semester%20Homework%207%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0A%40app.cell%0Adef%20_(mo)%3A%0A%20%20%20%20mo.md(r%22%22%22%0A%20%20%20%20%23%23%20Problem%201%0A%0A%20%20%20%20**Use%20homotopy%20groups%20to%20show%20there%20is%20no%20retraction%20%24%5Cmathbb%7BR%7DP%5En%20%5Cto%20%5Cmathbb%7BR%7DP%5Ek%24%20if%20%24n%3Ek%3E0%24**.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Suppose%20there%20is%20such%20retraction%2C%20then%20the%20induced%20homomorphism%20of%20inclusion%20%24i_*%3A%20%5Cpi_k(%5Cmathbb%7BR%7DP%5Ek)%20%5Cto%20%5Cpi_k(%5Cmathbb%7BR%7DP%5En)%24%20is%20injective.%20We%20shall%20show%20this%20is%20impossible.%0A%0A%20%20%20%20If%20%24k%20%3D%201%24%2C%20then%20%24%5Cpi_1(%5Cmathbb%7BR%7DP%5E1)%20%5Ccong%20%5Cpi_1(S%5E1)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%20but%20%24%5Cpi_1(%5Cmathbb%7BR%7DP%5En)%20%5Ccong%20%5Cmathbb%7BZ%7D_2%24%2C%20so%20there%20is%20no%20injective%20homomorphism%20between%20these.%0A%0A%20%20%20%20For%20%24k%20%5Cge%202%24%2C%20since%20%24S%5Em%24%20is%20a%20covering%20space%20for%20%24%5Cmathbb%7BR%7DP%5Em%24%2C%20we%20have%20that%20%24%5Cpi_k(S%5Em)%20%5Ccong%20%5Cpi_k(%5Cmathbb%7BR%7DP%5Em)%24.%20We%20know%20that%20%24%5Cpi_k(S%5Ek)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%2C%20so%20we%20have%20that%20%24%5Cpi_k(%5Cmathbb%7BR%7DP%5Ek)%20%5Ccong%20%5Cmathbb%7BZ%7D%24.%20We%20also%20know%20that%20for%20%24k%20%3C%20m%24%2C%20%24%5Cpi_k(S%5Em)%20%3D%200%24%2C%20thus%20%24%5Cpi_k(%5Cmathbb%7BR%7DP%5En)%20%3D%200%24.%20So%20again%20there%20cannot%20be%20an%20injective%20homomorphism%20from%20%24%5Cmathbb%7BZ%7D%20%5Cto%200%24%20and%20we%20get%20the%20desired%20result.%0A%0A%20%20%20%20%23%23%20Problem%2014%0A%0A%20%20%20%20**If%20an%20%24n%24-dimensional%20CW-complex%20%24X%24%20contains%20a%20sub-complex%20%24Y%24%20homotopy%20equivalent%20to%20%24S%5En%24%2C%20show%20that%20the%20map%20%24%5Cpi_n(Y)%20%5Cto%20%5Cpi_n(X)%24%20induced%20by%20inclusion%20is%20injective**.%20*Hint%3A%20Use%20the%20Hurewicz%20homomorphism*.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20If%20%24n%20%3D%200%24%2C%20then%20%24X%24%2C%20%24Y%24%2C%20%24%5Cpi_0(X)%24%20and%20%24%5Cpi_0(Y)%24%20are%20all%20discrete%20sets%20and%20the%20statement%20is%20trivial.%20If%20%24n%20%3D%201%24%2C%20then%20%24X%24%20is%20a%20graph%2C%20%24Y%24%20is%20homotopy%20equivalent%20to%20a%20circle.%20The%20image%20of%20a%20non-trivial%20loop%20in%20%24Y%24%20under%20the%20inclusion%20map%20is%20still%20a%20non-trivial%20loop%20in%20%24X%24%2C%20so%20the%20map%20%24%5Cpi_1(Y)%20%5Cto%20%5Cpi_1(X)%24%20is%20in%20fact%20also%20an%20inclusion%20and%20thus%20injective.%0A%0A%20%20%20%20Now%20suppose%20%24n%20%5Cge%202%24.%20Since%20%24Y%20%5Csimeq%20S%5En%24%2C%20%24H_n(Y)%20%5Ccong%20%5Cmathbb%7BZ%7D%24%20generated%20by%20certain%20%24%5B%5Calpha%5D%24.%20It%20is%20also%20an%20%24n%24-cycle%20in%20%24X%24%20because%20%24Y%24%20is%20a%20sub-complex%2C%20and%20since%20%24X%24%20is%20%24n%24-dimensional%2C%20there%20is%20no%20%24n%24-boundary%20and%20thus%20%24%5B%5Calpha%5D%24%20is%20non-trivial%20in%20%24H_n(X)%24.%20In%20particular%2C%20%24H_n(Y)%20%5Cto%20H_n(X)%24%20induced%20by%20inclusion%20is%20injective.%20Consider%20the%20Hurewicz%20homomorphism%2C%20then%20%24%5Cpi_n(Y)%20%5Cxrightarrow%7Bh%7D%20H_n(Y)%20%5Cxrightarrow%7Bi_*%7D%20H_n(X)%24%20is%20injective%20because%20%24h%24%20is%20an%20isomorphism%20for%20%24n%20%5Cge%202%24%20(this%20is%20due%20to%20the%20Hurewicz%20Theorem%2C%20it%20applies%20because%20%24%5Cpi_0(Y)%20%3D%20%5Cdots%20%3D%20%5Cpi_%7Bn-1%7D(Y)%20%3D%200%24%20as%20it%20is%20homotopy%20equivalent%20to%20%24S%5En%24).%20But%20this%20is%20the%20same%20map%20as%20%24%5Cpi_n(Y)%20%5Cxrightarrow%7Bi_*%7D%20%5Cpi_n(X)%20%5Cxrightarrow%7Bh%7D%20H_n(X)%24%20because%20both%20maps%20the%20generator%20%24%5Bf%5D%24%20of%20%24%5Cpi_n(Y)%24%20to%20%24%5Bi%5Ccirc%20f%20%5Ccirc%20%5Calpha%5D%24%20where%20%24%5Calpha%24%20is%20the%20fundamental%20class%20of%20%24S%5En%24.%0A%0A%20%20%20%20Since%20%24f%5Ccirc%20g%24%20injective%20implies%20%24g%24%20injective%2C%20we%20have%20that%20the%20part%20%24%5Cpi_n(Y)%20%5Cto%20%5Cpi_n(X)%24%20is%20injective.%0A%0A%20%20%20%20%23%23%20Problem%2015%0A%0A%20%20%20%20**Show%20that%20a%20closed%20simply-connected%20%243%24-manifold%20is%20homotopy%20equivalent%20to%20%24S%5E3%24**.%20*Hint%3A%20Use%20Poincar%C3%A9%20duality%2C%20and%20also%20the%20fact%20that%20closed%20manifolds%20are%20homotopy%20equivalent%20to%20CW-complexes%2C%20from%20Corollary%20A.12%20in%20the%20Appendix.%20The%20stronger%20statement%20that%20a%20closed%20simply-connected%20%243%24-manifold%20is%20homeomorphic%20to%20%24S%5E3%24%20is%20the%20Poincar%C3%A9%20conjecture%2C%20finally%20proved%20by%20Perelman.%20The%20higher-dimensional%20analog%2C%20that%20a%20closed%20%24n%24-manifold%20homotopy%20equivalent%20to%20%24S%5En%24%20is%20homeomorphic%20to%20%24S%5En%24%2C%20had%20been%20proved%20earlier%20for%20all%20%24n%20%5Cge%204%24*.%0A%0A%20%20%20%20Proof%0A%0A%20%20%20%20Let%20us%20call%20this%20manifold%20%24M%24%2C%20by%20the%20hint%2C%20we%20can%20assume%20it%20is%20instead%20a%20CW-complex.%20Since%20%24M%24%20is%20simply-connected%2C%20%24%5Cpi_1(M)%20%3D%200%24%20thus%20%24H_1(M)%20%3D%200%24%20by%20Hurewicz.%20Simply-connectedness%20also%20implies%20that%20%24M%24%20is%20orientable%2C%20thus%20%24H_3(M)%20%5Ccong%20%5Cmathbb%7BZ%7D%24.%20Being%20orientable%20means%20Poincar%C3%A9%20duality%20applies%2C%20so%20%24H_2(M)%20%3D%20H%5E1(M)%24%20which%20then%20by%20UCT%20%24%3D%20%5Ctext%7BExt%7D(H_0(M)%2C%5Cmathbb%7BZ%7D)%20%5Coplus%20%5Ctext%7BHom%7D(H_1(M)%2C%5Cmathbb%7BZ%7D)%24%20%24%3D%200%20%5Coplus%200%20%3D%200%24.%20The%20higher%20homology%20groups%20are%20of%20course%20trivial.%0A%0A%20%20%20%20By%20Hurewicz%20Theorem%2C%20%24%5Cpi_3(M)%20%5Ccong%20%5Cmathbb%7BZ%7D%24.%20Let%20%24f%3A%20S%5E3%20%5Cto%20M%24%20(understand%20as%20a%20spheroid)%20be%20the%20generator%20of%20%24%5Cpi_3(M)%24%2C%20then%20its%20induced%20homomorphism%20is%20the%20isomorphism%20between%20%24H_3(S%5E3)%20%5Cto%20H_3(M)%24%2C%20because%20the%20image%20of%20%24f%24%20under%20the%20Hurewicz%20homomorphism%20is%20the%20generator%20of%20%24H_3(M)%24%2C%20which%20by%20construction%20is%20also%20the%20image%20of%20the%20fundamental%20class%20in%20%24H_3(S%5E3)%24.%0A%0A%20%20%20%20For%20%24n%24%20that%20is%20not%20%240%24%20or%20%243%24%2C%20%24f%24%20of%20course%20induces%20isomorphisms%20from%20%24H_n(S%5E3)%20%5Cto%20H_n(M)%24%20because%20they%20are%20all%20just%20trivial.%20The%20isomorphism%20on%20the%20%240%24-th%20homology%20is%20also%20straight-forward.%20Thus%20%24f%3A%20S%5E3%20%5Cto%20M%24%20induces%20isomorphisms%20on%20all%20homology%20groups%2C%20thus%20by%20Whitehead%20Theorem%2C%20%24S%5E3%24%20and%20%24M%24%20are%20homotopy%20equivalent.%0A%20%20%20%20%22%22%22)%0A%20%20%20%20return%0A%0A%0Aif%20__name__%20%3D%3D%20%22__main__%22%3A%0A%20%20%20%20app.run()%0A