0.1.

Construct an explicit deformation retraction of the torus with one point deleted onto a graph
consisting of two circles intersecting in a point, namely, longitude and meridian circles of the torus.
Solution. T 2 can be regarded as the quotient space of R2 under the identification (x + n, y) ∼ (x, y) ∼
(x, y + m) for n, m ∈ Z. We will describe the points on T 2 through the points R2 using this quotient map.
Suppose the removed point from T 2 is the point (0, 0), which we will write  as ∗. If x ∈ R let [x] be the closest
integer to x. To remove the ambiguity at half-integers, we choose define n + 21 to be equal to n. Also define
x to be x − [x]. Now we define the following homotopy on T 2 − {∗}:
 
1
ft (x, y) = (x, y) + t − 1 (x, y).
2 max(|x|, |y|)
This homotopy is well-defined, since ft (x + n, y + m) = ft (x, y) + (n, m). It is straightforward to check that
the function is continuous; however, we are not going to delve into that hairy business. We want to show that
this is a deformation retraction to points (x, y), such that either x or y is a half-integer. Call this subspace
H. We will be done with this, since H is a wedge of two circles. One can easily verify that f0 = 1 and the
image of f1 is within H. Note that if (x, y) is within H, then max(|x|, |y|) = 12 ; therefore, ft |H = 1.
0.2. Construct an explicit deformation retraction of Rn − {0} onto S n−1 .
Solution. The following homotopy does the work:
 
t
f (x, t) = x + (1 − t) .
|x|

0.3. (a) Show that the composition of homotopy equivalences X → Y and Y → Z is homotopy equivalence
X → Z. Deduce that homotopy equivalence is an equivalence relation.
(b) Show that the relation of homotopy among maps X → Y is an equivalence relation.
(c) Show that a map homotopic to a homotopy equivalence is a homotopy equivalence.
Solution. (a) First we show that if we have ν : X → Y and µ, λ : Y → Z, such that µ ' λ, then µν ' λν.
If the homotopy from µ to λ is H, then the map H ◦ (ν × 1) is a homotopy from λν to µν. The same statement
holds for left multiplications by continuous functions. Suppose now that ψ : X → Y and ϕ : Y → Z are
homotopy equivalences, with inverses ψe and ϕ e ' ϕ1ϕ
e respectively. Now compute ϕψ ψeϕ e = ϕϕe ' 1. Similarly,
ψeϕϕψ
e ' 1.
(b) A map ψ : X → Y is clearly equivalent to itself via the homotopy ψ × 1. Suppose now that ψ is
homotopic to ϕ via homotopy H : X × I → Y . Define r : I → I to be the function r(x) = 1 − x. The map
H ◦ (1 × r) defines a homotopy from ϕ to ψ. Finally suppose that ψ is homotopic to ϕ via H1 , and ϕ to ν
via H2 . We define the following homotopy:
(
H1 (x, 2t) if t ≤ 21
H(x, t) = .
H2 (x, 2t − 1) if t ≥ 21

The function is continuous, and it defines a homotopy from ψ to ν.

(c) Suppose ψ : X → Y is a homotopy equivalence with an inverse ψ,
e and suppose ϕ ' ψ. Then
ϕψ ' ψ ψ ' 1, and in a similar way ψϕ ' 1.
e e e
0.4. A deformation retraction in the weak sense of a space X to a subspace A is a homotopy ft : X → X
such that f0 = 1, f1 (X) ⊂ A, and ft (A) ⊂ A for all t. Show that if X deformation retracts to A in this weak
sense, then the inclusion A ,→ X is a homotopy equivalence.
Solution. Let i denote the inclusion map and ρ denote the restriction of the range of f1 to A. In fact, ρ
is the homotopy inverse of i. From the conditions that we are given from the problem, we can restrict the
domain and range of ft to A, to obtain a homotopy gt on A. One can rather easily see that g0 = 1, since
f0 = 1. It also follows from the definition, that g1 = ρi. Therefore, ρi ' 1. On the other hand, iρ is the same
as f1 ; hence, iρ ' 1. Thus, i is a homotopy equivalence.
0.5. Show that if a space X deformation retracts to a point x ∈ X, then for each neighborhood U of x in
X there exists a neighborhood V ⊂ U of x such that the inclusion map V ,→ U is nullhomotopic.
Solution. Let the deformation retraction to the point x, be f : X × I → X. Consider the open set f −1 (U ).
Notice that {x}×I ⊂ f −1 (U ). One can use the tube lemma to show that there exists an open set V containing

1
x, such that V × I ⊂ f −1 (U ). Let ϕ : V × I → U be the restriction of f to domain V × I and range U . Since
f0 = 1, we have that g0 is the inclusion map of V into U . f1 on the other hand is a constant map, implying
that g1 is also a constant map. Thus, the inclusion map is nullhomotopic.
0.6. (a) Let X be the subspace R2 consisting of the horizontal segment [0, 1] × {0} together with all the
vertical segments {r} × [0, 1 − r] for r, a rational number in [0, 1]. Show that X deformation retracts to any
point in the segment [0, 1] × {0}, but not to any other point.

(b) Let Y be the subspace of R2 that is the union of an infinite number of copies of X arranged as in the
figure below. Show that Y is contractible but does not deformation retract onto any point.

(c) Let Z be the zigzag subspace of Y homeomorphic to R indicated by the heavier line. Show there is a
deformation retraction in the weak sense of Y onto Z, but no true deformation retraction.
Solution. (a) The deformation retraction to [0, 1] × {0} is the homotopy H(x, y, t) = (x, y(1 − t)). Then
we can go ahead and deformation retract [0, 1] × {0} to any of its points. If concatenate these deformation
retraction, we can get a deformation retraction to any point on [0, 1] × {0}. For the points that are not in
the bottom interval, we can observe that for exists disconnected neighborhood U , any neighborhood V ⊂ U
of x is not contained in a single component of U . Then V ,→ U cannot be nullhomotopic. Therefore, no
deformation retraction can be performed according to exercise 0.5.
(b) Let us denote the middle zigzag by Z. We can construct a homotopy on Y , such that it is deformation
retraction in the weak sense onto Z. Instead of giving an explicit formula, we will describe the homotopy
in terms of the motion of individual points. Each point moves with velocity of magnitude 1 at each time
instance. If the point is at a particular point in time is on Z, then it moves along the Z in the right direction.
If the point is not on the Z, then it moves towards Z along the brush-line. This defines a continuous√function
at any point in time, since the distance between two points at any time instance cannot exceed 2 of the
original distance. In addition, the points originally on the Z remain on the Z. By the end of the motion,
everything ends up being on the Z. Thus, Y has the same homotopy type as Z, and therefore, is contractible.
Y does not deformation retract to any point, since every point x ∈ Y we can choose a disconnected
neighborhood U containing x, such that any neighborhood of x within U is not contained within a single
component of U . This is true, since it implies that no neighborhood of x within U can have a nullhomopotic
inclusion onto U .
(c) We have constructed the deformation retract in the weak sense. We are left to show that there is no true
deformation retract. If there were such a retract we would have been able first to deformation retract to Z and
then to any point of Z. We have shown, that this is impossible in part (b).

0.7. Fill in the details in the following construction from [Edwards

1999] of a compact space Y ⊂ R3 with the same properties as the space Y
in exercise 6, that is, Y is contractible but does not deformation retract to
any point. To begin, let X be the union of an infinite sequence of cones on
Cantor set arranged end-to-end as in the figure. Next, form the one-point
compactification of X × R. This embeds in R3 as a closed disk with curved
‘fins’ attached along circular arcs, and with the one-point compactification
of X as a cross-sectional slice. The desired space Y is then obtained from
this subspace of R3 by wrapping one more cone on the Cantor set around X
the boundary of the disk. Y
Solution. It is easy to understand that the space cannot deformation retract to a point. The argument
is exactly the same as in the previous problem. We are left to show that Y is contractible. As we can see
X embeds in R2 . It is not closed in R2 ; however, X is homeomorphic to one-point compactification of X. It

2
is not difficult to construct a weak deformation retraction from X to J, the closed vertical interval that it
contains. Collapse the first cone to its cone-vertex, then successively collapse n-th cone to the interval that is
the intersection of J with (n − 1)-th cone. For the remaining point, let the deformation remain constant. In
this situation one has to be careful not to mess up the continuity of the map; however, it is quite easy to do.
Notice that the deformation retraction leaves ∂J unshifted. Let ∗ be the point of Y from which the outer fins
come out from, and let Z be the one-point compactification of X × R. If take a plane P in R3 that passes
through ∗, then P ∩ Z will be a scaling of X. We weak deformation retract this intersection for each plane,
so that the resulting function is continuous. The resulting deformation is a weak deformation retraction of
Z onto a closed disk D. Furthermore, on ∂D the weak deformation retraction is constant. Therefore, this
extends to a weak deformation retraction from Y to D ∪ O, where O is the outer cone on Cantor set of Y .
D ∪ O, on the other hand, weak deformation retracts to D in the following way: the points in D ∪ O move
at a constant angular velocity so that after time 1, they make a 2π angle revolution. Thus, Y ' D ∪ O ' D,
and the latter is contractible.
0.8. For n > 2, construct an n-room analog of the house with two rooms.
Solution. The following picture more less demonstrates construction for the n = 3 case.

∪ ∪

It is easy to generalize the construction for arbitrary n. We first take n + 1 rectangles of equal size and
vertically equally spaced as in the picture. Then we ‘punch’ n square-shaped holes on the top rectangle, n − 1
holes on the next one, and so on, until we exhaust all the rectangles. Notice that the last one will not be
‘punched’. This must be done in such a way so that the squares line up on top of one another, again as in the
picture. Then we take boxes that miss top and bottom faces, and so that the missing face exactly matches
the dimensions of the ‘punched’ hole and the hight equals the distance between rectangles. Then we glue
n − 1 such boxes and ‘glue’ them in a single row. Then right underneath this row, we ‘glue’ n − 2 boxes in
a single row, so that the leftmost box of the top row does not have a box underneath. Continue in a similar
way. Once done with this construction, also attach copies of the leftmost face as it is shown in the picture.
Finally, add the caracass and glue everything together. This space will be contractible and it will contain n
rooms.
0.9. Show that the retract of a contractible space is contractible.
Solution. Suppose X is contractible, A ⊂ X and r : A → X is a retraction. Since X is contractible,
then there exist a homotopy H : X × I → X from 1 to some constant map ν. The following function
K = r ◦ H ◦ (i × 1) maps A × I to A continuously; therefore, it is a homotopy. Notice that K0 = r ◦ i = 1
and K1 = r ◦ ν ◦ i is constant, since ν is constant. Thus, A is contractible.
0.10. Show that a space X is contractible iff every f : X → Y , for arbitrary Y , is nullhomotopic. Similarly,
show X is contractible iff every map f : Y → X is nullhomotopic.
Solution. If arbitrary f : X → Y is nullhomotopic arbitrary f : Y → X is nullhomotopic, then, in
particular, 1 : X → X is nullhomotopic, and for that reason X is contractible. Now suppose X is contractible.
Therefore, there is a homotopy H : X × I → X from 1 to a constant map. If f : X → Y is a map, then f ◦ H
defines a nullhomotopy for f . Similarly, for f : Y → X, H ◦ (f × 1) a nullhomotopy for f .
0.11. Show that f : X → Y is a homotopy equivalence if there exist maps g, h : Y → X, such that f g ' 1
and hf ' 1. More generally, f is homotopy equivalence if f g and hf are homotopy equivalences.
Solution. The solution to the first part of the problem is fairly straightforward gf ' hf gf ' hf ' 1.
The second is also straightforward if we clarify the conditions. f g is a homotopy equivalence, then it means
that there exists a mapping s : Y → Y , such that sf g ' 1 and f gs ' 1. Similarly, there exist a mapping

3
r : X → X, such that hf r ' 1 and rhf ' 1. However, from the first part we know that f gs ' 1 and rhf ' 1
imply that f is a homotopy equivalence.
0.12. Show that a homotopy equivalence f : X → Y induces a bijection between the set of path-
components of X and the set of path-components of Y , and that f restricts to a homotopy equivalence from
each path-component of X to the corresponding path-component of Y . Prove also the corresponding statement
with components instead of path-components. Deduce from this that if the components and path-components
of a space coincide, then the same is true for any homotopy equivalent space.
Solution. Let Π(X) denote the set of path components of X. Suppose we are given a continuous map
ν : X → Y . Define a function ν∗ : Π(X) → Π(Y ), so that it maps P ∈ Π(X) to the path-component in Y
containing ν(P ). One can also easily show that the induced functions respect composition, i.e. (νκ)∗ = ν∗ κ∗ ,
and 1∗ = 1. Suppose we are given two maps ν, µ : X → Y , so that ν ' µ. We claim that ν∗ = µ∗ . Let
H : X × I → Y be a homotopy between ν and µ. Suppose P ∈ Π(X). The subspace P × I of X × I is
path-connected. Notice that ν(P ) ⊂ H(P × I) and µ(P ) ⊂ H(P × I); therefore, ν(P ) and µ(P ) map to the
same path-component. Now suppose that f : X → Y is homotopy equivalence with an inverse g : Y → X.
Then, f∗ g∗ = (f g)∗ = 1∗ = 1 and g∗ f∗ = 1. Thus, f∗ is a bijection. Notice that H the path-connected
subspace P × I to a single component of Y ; thus, restriction of H|P ×I can be regarded as a homotopy
between path-components. If Q = f∗ (P ), then we can regard f |P : P → Q and g|Q : Q → P . Then we
g|Q f |P = (gf )|P ' 1; similarly the other direction follows. For the component exact same arguments work.
0.13. Show that any two deformation retractions rt0 and rt1 of a space X onto a subspace A can be joined
by a continuous family of deformation retractions rts , 0 ≤ s ≤ 1, of X onto A, where continuity means that
the map X × I × I → X sending (x, s, t) to rts (x) is continuous.
Solution. The construction consists of two parts. In first part we construct a family of functions connecting
rt0 to rt1 . The function is defined from X × I × I → X:

0
rt−2s
 if t − 2s ≥ 0
s 1
pt = rt+2s−2 if t + 2s ≥ 2 .

1 otherwise


One can easily check that the function is well-defined and continuous. In the second step we ‘deformation
retractify’ pst by composing it with rt0 . At the end of the day we have

rts = rt0 pst .

It is easy to verify that all the conditions are satisfied.

0.14. Given positive integers v, e and f satisfying v − e + f = 2, construct a cell structure on S 2 having
v 0-cells, e 1-cells, and f 2-cells.
Solution. Suppose we have constructed a cell-structure for a sphere so that there are (v, e, f ) number
of cells. We show that it will be possible to give a cell-structure so that it contains (v + 1, e + 1, f ) and
(v, e + 1, f + 1) cells. For the first case, we take an arbitrary 1-cell, e1 ; it exists, since e > 0. Then add a
0-cell at the midpoint of e1 , replace e1 with two 1-cells, where the first one maps to (0, 12 ) portion of e1 and
the other to ( 12 , 1) portion. For the second case, pick a 2-cell e2 . Add a 1-cell by sending e1 to a diameter of
e2 . Replace e2 by two 2-cell, where each is identified with the complement of that diameter in e2 . It is easy
to give a cell-structure of (1, 1, 2) cells and (2, 1, 1) cells for S 2 .
0.15. Enumerate all the subcomplexes of S ∞ , with the cell structure described in this section, having two
cells in each dimension.
Solution. Suppose that S ⊂ S ∞ is a subcomplex. If S contains a n-cell, say en , then S n−1 = ∂en ⊂ S.
Therefore, if dim S = ∞, then S n ⊂ S for any n ∈ N, which implies that S = S ∞ . Now suppose dim S = n.
Suppose it contains one n-cell, en . Then, en ⊂ S, contains all the cells of lower dimensions, and in itself is a
subcomplex. If S contains two n-cells, then S n = S.
0.16. Show that S ∞ is contractible.
Solution. Let ∗ be one of the 0-cells of S ∞ . There is a homotopy rn : S n × I → S n collapsing S n−1 to ∗,
without move ∗ at any point in time. Since (S ∞ , S n ) is a homotopy extension pair, then there is a homotopy

4
ren : S ∞ × I → S ∞ , which collapses S n−1 to ∗. Concatenate these homotopies with one another. Define
r : S ∞ × I → S ∞ , so that r|[1− 1 ,1− 1n ] = ren ◦ re1n−1 ◦ ... ◦ re11 and r1 sends everything to ∗. If consider
2n−1 2
the restriction of r any S n , we will see that it continuously deforms S n to ∗ and after that remains constant.
Therefore, r is continuous. Notice that r defines a deformation retraction to a point.
L.1. Show that the mapping cylinder of every map f : S 1 → S 1 is a CW complex.
Solution. Start with two 0-cells and then attach to them three 1-cells. Construct the complex depicted
below. Do it in a way so that 0-cell in the blue circle maps via f to 0-cell in the red circle.

We attach a 2-cell. Here it will be convenient for us to think of the 2-cell as a closed square I 2 . Map both
intervals in ∂I × I to the interval in 1-skeleton, so that ∂I × {1} ends up in the blue circle and ∂I × {0} ends
up in the red circle. Then attach {1} × I to the blue circle, so that it agrees with previous map and wraps
around the circle once, i.e. is injective everywhere except on the boundary. Then attach {0} × I to the red
circle, by attaching (0, x) to the image under f of the attachment point of (1, x). This will end up being the
mapping cylinder of f .
0.17. Construct a 2-dimensional cell complex that contains both an annulus S 1 × I and a Möbius band
as deformation retracts.
Solution. Let σ : S 1 → S 1 is a double-covering map of S 1 , then Mσ is the Möbius band. M1 on the hand
is the annulus. Let M be the attachment of Mσ to M1 by the identification of the base circles. By collapsing,
M1 to its base circle, we get a deformation retraction to Mσ ; doing it the other way, we get deformation
retraction to M1 .
0.18. Show that S 1 ∗ S 1 = S 3 , and more generally S n ∗ S m = S n+m+1 .
n+1 m+1
Solution.

Defineπϕthe

following functionnf : Rm × R × I −→ Rn+m+2 in following way f (x, y, ϕ) =
πϕ n+m+1
x cos 2 , y sin 2 . Notice that f (S × S × I) ⊂ S . Let from this point on consider f the
domain and range restricted to these subspaces. Clearly, f is continuous. Since the domain is compact and
the range is Hausdorff, then f is closed. f is also surjective. Therefore, S n+m+1 is a quotient

space of S

n ×
m πϕ πϕ
 × I. Now
S  let us see
what the equivalence relation established by f is. Suppose x cos 2 , y sin 2 =
e cos π2ϕe , y
x e sin π2ϕe . It is easy to see that ϕ = ϕ.
e If ϕ 6= 0, 1, then x = x
e and y = y
e . However, ϕ = 0,
then x = x e and y, y
e can be arbitrary. The same happens in the other direction for ϕ = 1. However, this
precisely make S n+m+1 into S n ∗ S m .
0.19. Show that the space obtained from S 2 by attaching n 2-cells along any collection of n circles in S 2
is homotopy equivalent to the wedge sum of n + 1 2-spheres.
Solution. The space the we have can be written as S 2 f X, where X = Dk2 and f : ∂Dk2 → S 2 , which
` ` `
sends the boundaries of the disks to the respective circle. A union of finite collection of spheres cannot be the
whole space S 2 ; therefore, this union is contained in`S 2 − {∗}, for some ∗ ∈ S 2 . This subspace is contractible,
and for that reason f is nullhomotopic. Thus, S 2 f X is homotopy equivalent to S 2 g X, where g sends
`
the boundaries of the disks to a single point. However, the last space is exactly a wedge of spheres.
0.20. Show that the subspace X ⊂ R3 formed by the Klein bottle intersecting itself in a circle, as shown
in the figure, is homotopy equivalent to S 1 ∨ S 1 ∨ S 2 .

Solution. The intersection disk is contractible; therefore, we can collapse the disk to a point without changing
the homotopy type. This new space can also be obtain by collapsing the blue lines of the space depicted

5
below (sphere with two intervals attached to it). All one needs to do is to collapse the blue intervals. On the
other hand we can collapse the gray interval instead, and we instant obtain S 1 ∨ S 1 ∨ S 2 . Since, the collapsed
subspaces were contractible then all the spaces had the same homotopy type.

0.21 If X is a connected space that is a union of finite number of 2-spheres, any two of which intersect in
at most one point, show that X is homotopy equivalent to a wedge sum of S 1 ’s and S 2 ’s.
Solution.