Time flies

Geometry Level 2

Can the hour, minute, and second hands—when extended to the circumference—ever cut a properly functioning circular clock into three equal areas?

Yes No

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

5 solutions

Antoine G
Jun 19, 2018

To simplify somehow the problem let us make 60 ticks for all hands. This said the position of a hand can be modeled by the group R / 60 Z \mathbb{R} / 60 \mathbb{Z} (or in other words we are looking at the reals numbers modulo 60; see Details below if you are not familiar with this). This is required in order to deal with continuously moving hands (the second hand could be halfway between two seconds).

The actual possible positions for the hands are given by the triplet ( t , t 60 , t 720 ) (t, \frac{t}{60}, \tfrac{t}{720} ) . This is because the minute hand turns 60 times slower than the second hand, and the hour hand turns 12 times slower than the minute hand (so 720 times slower than the second hand). In order that the hands split the clock in three equal thirds, it must hold that the difference between any two coordinates is ± 20 mod 60 \pm 20 \text{ mod } 60 (because one third of the clock is 20 units). This means ( ) t t 60 = ± 20 mod 60 ( ) t 60 t 720 = ± 20 mod 60 \begin{array}{rrll} (*) & t - \dfrac{t}{60} & = \pm 20 & \text{mod } 60 \\ (**) & \dfrac{t}{60} - \dfrac{t}{720} & = \pm 20 & \text{mod } 60 \\ \end{array} However note that ( ) : t 60 t 720 = ± 20 mod 60 720 t 60 720 t 720 = 720 ± 20 mod 60 12 t t = 0 mod 60 11 t = 0 mod 60 \begin{array}{rrll} (**): & \dfrac{t}{60} - \dfrac{t}{720} & = \pm 20 & \text{mod } 60 \\ \implies & 720 \cdot \dfrac{t}{60} - 720 \cdot \tfrac{t}{720} & = 720 \cdot \pm 20 & \text{mod } 60 \\ \implies & 12 \cdot t - t & = 0 & \text{mod } 60 \\ \implies & 11 \cdot t & = 0 & \text{mod } 60 \\ \end{array} Call this last equation ( ) (***) . Next ( ) : t t 60 = ± 20 mod 60 60 t 60 t 60 = 60 ± 20 mod 60 60 t t = 0 mod 60 59 t = 0 mod 60 \begin{array}{rrll} (*): & t - \dfrac{t}{60} & = \pm 20 & \text{mod } 60 \\ \implies & 60 \cdot t - 60 \cdot \dfrac{t}{60} & = 60 \cdot \pm 20 & \text{mod } 60 \\ \implies & 60 \cdot t - t & = 0 & \text{mod } 60 \\ \implies & 59 \cdot t & = 0 & \text{mod } 60 \\ \end{array} Call this last equation ( ) (\bigstar) . Now it's not too hard to show ( ) (***) and \bigstar lead to a contradiction. ( ) : 59 t = 0 mod 60 11 t + 11 t + 11 t + 11 t + 11 t + 4 t = 0 mod 60 ( ) 0 + 0 + 0 + 0 + 0 + 4 t = 0 mod 60 4 t = 0 mod 60 \begin{array}{rrll} (\bigstar): & 59 \cdot t & = 0 & \text{mod } 60 \\ \implies & 11 \cdot t +11 \cdot t + 11 \cdot t+ 11 \cdot t+ 11 \cdot t + 4 \cdot t & = 0 & \text{mod } 60 \\ \overset{(***)}{\implies} & 0+0+0+0+0+ 4\cdot t & = 0 & \text{mod } 60 \\ \implies & 4 \cdot t & = 0 & \text{mod } 60 \\ \end{array} Now repeat the same trick 11 t = 4 t + 4 t + 3 t = 0 mod 60 11 t = 4t + 4t + 3t = 0 \text{ mod } 60 and hence 3 t = 0 mod 60 3 t = 0 \text{ mod } 60 . But then 4 t = 3 t + t = 0 mod 60 4 t = 3t + t = 0 \text{ mod } 60 so t = 0 mod 60 t = 0 \text{ mod } 60 .

So t t is an integer(!) and a multiple of 60. We could go on with the equations, but the contradiction is clear. This means the second hand is at the origin (mark 12 on the clock). Then the minute hand must be either at 20 or 40 minutes exactly (marks 4 or 8; in order to make the proper angle). But in that case the hour hand cannot be at marks 4 or 8 (because the minute hand is not at the mark 12).

Hence ( ) (*) and ( ) (**) imply that the answer is No \fbox{No} .

For those who want to go on with equations

Once we have that t = 0 mod 60 t = 0 \text{ mod } 60 , ( ) (**) becomes, upon multiplication by 720: t 12 = 0 mod 60 \dfrac{t}{12} = 0 \text{ mod } 60 . This means that t t is an integer multiple of 720, say t = 720 n t = 720n . But then ( ) (*) reads 12 n = ± 20 mod 60 12 n = \pm 20 \text{ mod } 60 ( where n n is an integer ). The values of 12 n mod 60 12 n\text{ mod } 60 are 0 , 12 , 24 , 36 0,12,24,36 and 48 48 . So there are no solutions, and the answer is no.

Details on R / 60 Z \mathbb{R} / 60 \mathbb{Z}

The arithmetic of R mod 60 \mathbb{R} \text{ mod } 60 is essentially the same as the usual modulo arithmetic. For the clock, you can picture this as follows: take a real r r and add or subtract 60 until it lies in the (half-open) interval [ 0 , 60 ) [0,60) . For example 20 mod 60 -20 \text{ mod } 60 is represented by 40 40 . Addition works fine: 53 + 26 mod 60 = 79 mod 60 = 19 mod 60 53+26 \text{ mod } 60 = 79 \text{ mod } 60 = 19 \text{ mod } 60 . When manipulating equations, multiplication by integers is also unproblematic (because it's just repeated addition). However multiplication by reals which are not integers does not make sense when you are manipulating an equation. Note also that if r r is some real then 60 r mod 60 60 r \text{ mod } 60 could be different from 0 mod 60 0\text{ mod } 60 (for example if r = 1 60 r = \tfrac{1}{60} then 60 r = 1 60 r = 1 .

Note

Note that all allowed positions for the hands form a 3-dimensional space. However the set of correct positions is only one-dimensional. Lastly, the set of positions where the hands split the clock in equal thirds is also one-dimensional. This means that it would have been a big coincidence if the answer would have been yes, as in general two one-dimensional space inside a 3-dimensional space do not intersect.

paraphrase please..

drought RWS&S - 2 years, 11 months ago

Log in to reply

Could you please provide an explanation as to what you are pointing to and meaning when you say "paraphrase please" ?

Antoine G - 2 years, 11 months ago

Log in to reply

Actually i am a novice at modulo arithmetic.. just a vague remembrance of a lesson at school.. But the problem here and the your solution struck a chord as to the application of the modular concept.The way you solved the problem is ingenious , but a lot of stuff to comprehend for a mathematical novice like me. However, i do not want to visit the basics again myself and lose the gist of the actual .. Thats the reason why i asked you to paraphrase, that is if you are only willing to take the pain.. like you could point to another example of a same nature but a more simple one .. that does not have all these reciprocals and sufficiently illustrate the beauty of modular arithmetic.. Thanks

drought RWS&S - 2 years, 11 months ago

Log in to reply

@Drought Rws&S Perhaps a nice starter is

https://www.youtube.com/watch?v=KZT5hrYOERs

it also has to do with a clock problem...

But tell me if you'd like more!

Antoine G - 2 years, 11 months ago

Log in to reply

@Antoine G first of all, i would like to thank you. your advise has been the triggering point for me to watch various episodes of Pbs infinity series. Excellent videos. Extremely well designed and entertaining to the core. of course, i would like to know more from you.

drought RWS&S - 2 years, 8 months ago

Imagine you have another clock, which has the correct time at midnight but moves at triple speed. Then at the required time, all the hands on this clock will coincide. But we can show (see below) that that only happens if they are all pointing straight up, ergo the original clock must say either 12:00, 4:00, or 8:00. However, none of those times satisfy the requirements.

Proof of "If all the hands coincide then they are all pointing straight up": Let H be the distance that the hour hand has traveled since midnight (measured in whatever units you like). Then the distance that the minute hand has traveled is 12H, and therefore 11H is an integral number of times around the clock. Similarly, the distance that the second hand has traveled is 60x12H, and therefore 59x12H is an integral number of times around the clock. But there are integers A and B such that 11A+60x12B=1 (they can be found easily, or we can use the fact that 11 and 60x12 are relatively prime) and it follows that H is also an integral number of times around the clock.

Peter Byers - 2 years, 11 months ago

One can also note that when the angle between the minute and hour hands is a multiple of 6*, the second hand must lie on top of the minute hand.

Alex Burgess - 2 years, 11 months ago

Log in to reply

I think what I wrote is just a formalisation of that idea. (Altough I must confess that I'm not sure to understand what you mean when you say that "the minute and hour hands is a multiple of 6*" )

Antoine G - 2 years, 11 months ago

the hour hand turns 60 times slower than the minute hand

Although there are 60 minutes in an hour, the hour hand turns 12 times slower than the minute hand.

Peter Byers - 2 years, 11 months ago

Log in to reply

thanks! corrected...

Antoine G - 2 years, 11 months ago

Assuming all hands have same 60 positions (0 at the top), what about time 00:21:41 ? Hour hand will be at position 1, minute hand at 21 and second at 41 Why doesn't this combination satisfy requirements?

Maksym Mazur - 2 years, 11 months ago

Log in to reply

well, at 00:21:41 the second hand is at 41 and the minute hand is at 21 and 41/60 (because it already went 41/60 of the way to the next minute). So the angle between the minute and second hand is already incorrect. Hence your combination does not satisfy the requirements.

Antoine G - 2 years, 11 months ago

Don’t know how to post a solution, so I’ll just write it in a reply.

So there’s one cycle of every possible position of the hands of a clock every 12 hours. If we assume that a, b and c is the angle between every hand and the starting position (12:00, 0°), then in order to cut the clock in three equal portions one of the following equations should be true: (a+b+c)/3=a; (a+b+c)/3=b; (a+b+c)/3=c. Because the sum of the angles has to be 3 times the middle angle.

Now let x, y and z be the amount of times the hour, minute and second hands of the clock respectively do a cycle (360°).

Then a=360-360x; b=12x360-360y; c=720x360-360z.

Let’s put that in the previous equations.

(360-360x+4320-360y+259200-360z)/3 should equal either a, b or c.

Making it simple we get 120(733-x-y-z) equals either a, b or c.

By the time the hour hand makes it to 360°, the minute hand does it 12 times and the second hand does it 720 times. That means y=12x; z=720x. Whereas 0<x<1. Because when the hour hand is on 12:00, all the other hands are there too, and it also satisfies the equations (a+b+c)/3 equals either a, b or c. But the hour hand only appears there once a cycle, so we need to find a different solution.

Putting it all together we get: 120(733-733x)=360-360x; 733-733x=1-x; 732=732x; x=1.

The equation is true only for a full cycle. That means there is no solution.

Garegin Arakelian - 2 years, 11 months ago

Log in to reply

Here is one thing I do not understand in your method: you say "let x, y and z be the amount of times the hour, minute and second hands of the clock respectively do a cycle". Further down you write "By the time the hour hand makes it to 360°, the minute hand does it 12 times and the second hand does it 720 times. That means y=12x; z=720x."

The fact that "y=12x; z=720x" for a full 12 hours does not mean it always holds.

For example after exactly one minute, x=0, y=1 and z=0.

Antoine G - 2 years, 11 months ago

Log in to reply

After one minute x=1/720; y=1/60; z=1. One minute is one cycle of the second hand.

Garegin Arakelian - 2 years, 11 months ago

Log in to reply

@Garegin Arakelian ok, so x,y and z are not necessarily integers. Then if I put these values in :

" a=360-360x; b=12360-360y; c=720360-360z."

then I get unconventional angles (for example b = 12354°). This way of measuring angles is not coherent with a previous claim. Indeed, your claim that "in order to cut the clock in three equal portions one of the following equations should be true: (a+b+c)/3=a; (a+b+c)/3=b; (a+b+c)/3=c." works only if the angles are all between 0° and 360°.

Antoine G - 2 years, 11 months ago

I noted that there is a finite number of positions where the minute and second hand make a 60 degree angle not containing the hour hand and an uncountably infinite number of possible hour hand positions at that time, only one of which is correct, making it almost impossible.

Vilim Lendvaj - 2 years, 11 months ago

Log in to reply

Right, this is basically the dimension argument I am pointing out it my note. The set of positions of the hands is 3 dimensional (one dimension for each hand). However, the set which arise through a normal process (that is all hand start at 0 and their speed are is 1 dimensional (well it's a "time-line"). Next the positions where they cut the clock in three thirds is 1-dimensions (you can rotate all hands together to make a new position). The answer would be true if those two 1-dimensional sets intersect; but in 3-dimensional space, this is unlikely.

Antoine G - 2 years, 11 months ago

What about possible irrational values of t?

Kermit Rose - 2 years, 11 months ago

Log in to reply

They are also included in the above argument (real number contain irrational numbers). You can do arithmetic mod 60 with irrationals, you have to be careful which properties hold.

A common mistake is for example that, with integers ( a mod 60 ) ( b mod 60 ) = a b mod 60 (a \text{ mod } 60) \cdot (b \text{ mod } 60) = a \cdot b \text{ mod } 60 (i.e. you can first compute the residue mod 60 and then multiply). This is wrong with real number.

Perhaps a nice video on the topic is

https://www.youtube.com/watch?v=KZT5hrYOERs

it also has to do with a clock problem..

Antoine G - 2 years, 11 months ago

If we force the second hand to each second, why are we not also forcing the minute and hour hands? Therefore, 4:40:00 and 8:20:00 are easy solutions with probably plenty of others.

More interesting would be all three hands moving continuously.

Chris Arsenault - 2 years, 10 months ago

Log in to reply

As I tried to stress, I am not moving the second second per second. t t is a real number, not an integer . The fact that one consider modulo does not mean one deals with integer. That's why I wrote a paragraph Details about R / 60 Z \mathbb{R}/60 \mathbb{Z} .

If you are not satisfied with the explanations there, there are many other sources. See for example:

https://www.youtube.com/watch?v=KZT5hrYOERs

Antoine G - 2 years, 10 months ago
Amay Singh
Jun 25, 2018

In each hour the hour and minute hand make 60 degrees angle twice (with seconds hand pointing at 12). Now, the only condition required is that the second hand also makes an angle of 60 degrees with each hand which is like

But the angle between the minute and hour hand is a bit more or less than 120 degrees. Therefore, the given condition is impossible.

Couldn't we in theory get the desired result if the second hand was not on the 12? After all, the three hands do not move at the same rate.

Steven Gottlieb - 2 years, 11 months ago

Log in to reply

You are right that this explanation is flawed. However, it is not possible to split the the clock into thirds: set the rate that the( second hand moves - the rate that the hour hand moves) seconds = (difference between rates of the minute and second hands) seconds, and you will see that the radian displacements can never be the same (i.e. both equal 2pi/3 as they must to split the clock into thirds).

Romik Ghosh - 2 years, 11 months ago

Due to the symmetry of the clock and the constant ratios of hand speeds, if there is a solution, there is one where the second hand is vertical. If there aren't any solutions where the second hand is vertical, there aren't any solutions. This an example of reducing the dimensions of a problem by using symmetry. Using this argument, you only need to check two positions to see if they are valid clock times. Brilliant!

Chris Maitland - 2 years, 11 months ago

If we are using a ticking clock this is correct, but if we are using a continuously moving(i.e. motor driven with geardowns) clock hand, then the problem would seem to be solvable.

James Felling - 2 years, 11 months ago

Concise solution. My "method" was even briefer - it occurred to me that if such an arrangement were to be possible there would be a number of such arrangements (as with problems involving only two hands on a clock), and therefore it would be comparatively easy to find one. As this was set as an advanced level problem it had to be fundamentally difficult, so the answer had to be no.

Thomas Sutcliffe - 2 years, 11 months ago
Chew-Seong Cheong
Jun 23, 2018

Let the angular displacements starting at time t = 0 t=0 at 00:00 of the hour, minute, and second hands be θ h \theta_h , θ m \theta_m , and θ s \theta_s respectively in degrees. For 0 t < 24 0 \le t < 24 in hours, we have θ h ( t ) = 30 t m o d 360 \theta_h (t) = 30 t \bmod 360 , θ m ( t ) = 360 t m o d 360 \theta_m (t) = 360 t \bmod 360 , and θ s ( t ) = 21600 t m o d 360 \theta_s (t) = 21600 t \bmod 360 . Let us first consider when the hour and minute hands are 12 0 120^\circ apart. This happens when:

θ m ( t ) θ h ( t ) = 120 n where n is a positive integer not divisible by 3. 360 t 30 t = 120 n t n = 4 n 11 Since t < 24 , n 65 θ s ( t n ) = 86400 11 n m o d 360 = 3240 11 n m o d 360 \begin{aligned} \theta_m (t) - \theta_h (t) & = 120\color{#3D99F6}n & \small \color{#3D99F6} \text{where }n \text{ is a positive integer not divisible by 3.} \\ 360t - 30 t & = 120n \\ t_n & = \frac {4n}{11} & \small \color{#3D99F6} \text{Since }t < 24, \ n \le 65 \\ \implies \theta_s (t_n) & = \frac {86400}{11}n \bmod 360 \\ & = \frac {3240}{11}n \bmod 360 \end{aligned}

There are two cases.

Case 1: When n m o d 3 = 1 n \bmod 3 = 1 , θ m ( t n ) θ h ( t n ) = 120 \theta_m(t_n) - \theta_h (t_n) = 120 , we are checking if θ s ( t n ) θ h ( t n ) = 240 \theta_s (t_n) - \theta_h (t_n) = 240 . Let's see θ s ( t n ) θ h ( t n ) = 3240 11 n 120 11 n = 3120 11 n \theta_s (t_n) - \theta_h (t_n) = \frac {3240}{11}n - \frac {120}{11}n = \frac {3120}{11}n , which is only an integer when n n is a multiple of 11 and only two of these 22 , 55 m o d 3 = 1 22, 55 \bmod 3 = 1 . But 3120 × 2 m o d 360 = 3120 × 5 m o d 360 = 120 240 3120 \times 2 \bmod 360 = 3120 \times 5 \bmod 360 = 120 \ne 240 .

Case 2: When n m o d 3 = 2 n \bmod 3 = 2 , θ m ( t n ) θ h ( t n ) = 240 \theta_m(t_n) - \theta_h (t_n) = 240 , we are checking if θ s ( t n ) θ h ( t n ) = 120 \theta_s (t_n) - \theta_h (t_n) = 120 . Let's see θ s ( t n ) θ h ( t n ) = 3240 11 n 120 11 n = 3120 11 n \theta_s (t_n) - \theta_h (t_n) = \frac {3240}{11}n - \frac {120}{11}n = \frac {3120}{11}n , which is only an integer when n n is a multiple of 11 and only two of these 11 , 44 m o d 3 = 2 11, 44 \bmod 3 = 2 . But 3120 × 1 m o d 360 = 3120 × 4 m o d 360 = 240 120 3120 \times 1 \bmod 360 = 3120 \times 4 \bmod 360 = 240 \ne 120 .

No , there is never a time that the three hands of the clock divide the clock face into three equal areas.

Kelvin Hong
Jun 28, 2018

Assume that it is possible.

Let the angle made by second, minute and hour hand to be α = 2 π 60 t , β = 2 π 3600 t , γ = 2 π 43200 t \alpha=\dfrac{2\pi}{60}t,\beta=\dfrac{2\pi}{3600}t,\gamma=\dfrac{2\pi}{43200}t , then we will have { cos ( α β ) = 1 2 cos ( β γ ) = 1 2 \begin{cases}\cos(\alpha-\beta)=-\frac12\\\cos(\beta-\gamma)=-\frac12\end{cases} Also we have α β β γ ± 2 π 3 ( m o d 2 π ) \alpha-\beta\equiv\beta-\gamma\equiv\pm\frac{2\pi}{3}\pmod{2\pi} If the sign of α β ( m o d 2 π ) \alpha-\beta\pmod{2\pi} and β γ ( m o d 2 π ) \beta-\gamma\pmod{2\pi} are opposite, then will leads to α γ ( m o d 2 π ) \alpha\equiv\gamma\pmod{2\pi} which is contradiction. Then for some suitable integer k , h k,h we have { t ( 1 60 1 3600 ) = k ± 1 3 t ( 1 3600 1 43200 ) = h ± 1 3 \begin{cases}t\bigg(\frac1{60}-\frac1{3600}\bigg)=k\pm\frac13\\t\bigg(\frac1{3600}-\frac1{43200}\bigg)=h\pm\frac13\end{cases} Eliminate t t , we retrieve this equation: 3600 59 ( k ± 1 3 ) = 43200 11 ( h ± 1 3 ) \frac{3600}{59}(k\pm\frac13)=\frac{43200}{11}(h\pm\frac13) Which can be simplify to 1320 k ± 4400 = 849600 h ± 283200 1320k\pm4400=849600h\pm283200 But if we take modulo 3 we will see ± 2 0 ( m o d 3 ) \pm2\equiv0\pmod3 so this is a contradiction, this situation can NOT occur.

Michael Mendrin
Jun 25, 2018

Let's look at the time it takes for the second, minute, and hour hand to travel 30 30 degrees on the clock, where S S is the time in seconds:

S s e c = 30 5 S_{sec}=\dfrac{30}{5}
S m i n = 30 300 S_{min}=\dfrac{30}{300}
S h r = 30 3600 S_{hr}=\dfrac{30}{3600}

The condition that the minute and hour hand are exactly 120 120 degrees apart is

( 30 300 30 3600 ) S = 120 + 360 n \left( \dfrac{30}{300}-\dfrac{30}{3600} \right) S=120+360n

where n n is an integer, so that

S = ( 14400 11 ) ( 3 n + 1 ) S=\left( \dfrac{14400}{11} \right) (3n+1)

We then check to see if for any n = 1 n=1 to 11 11 the difference between

S m o d 60 1 60 ( S m o d 3600 ) S\; mod\;60 - \dfrac{1}{60}(S \;mod \;3600)

is exactly 20 20 seconds. The actual results are

( 27.2727 , 10.9091 , 10.9091 , 27.2727 , 5.45455 , 43.6364 , 21.8182 , \ 0. , 21.8182 , 43.6364 , 5.45455 ) (27.2727, -10.9091, 10.9091, -27.2727, -5.45455, -43.6364, -21.8182, \ 0., 21.8182, 43.6364, 5.45455)

so that although some are close to 20 20 , none of them are exactly 20 20 , so that the answer is "no".

This solution assumes that the clock is analog and all hands move continuously.

You can do modular arithmetic with reals. Have a look at

https://www.youtube.com/watch?v=KZT5hrYOERs

Antoine G - 2 years, 11 months ago

Sir i didnt get the symbol Mod(S,60) . plzz say about it

vijay mohangekar - 2 years, 11 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...