# Primes

Watch
Announcements

Report

#2

This isn't a rigorous mathematical proof, but more a logical narrowing-down. All primes are odd except for 2. No primes end in the number 5, except for 5, because anything else that ends in 5 is divisible by 5 and hence not prime. Any set of primes where each successive prime was 2 more than the last one would have to have last digits from among these, in order: 1, 3, 5, 7, 9. There is no way of doing that without including 5 as a last digit. Hence the

Nevermind, that doesn't work. You could still have a run of primes where the last digits went 7, 9, 1 or 9, 1, 3. I may still have generally been on the right track, though; perhaps there is some way of showing that if the last digits went 7, 9, 1 or 9, 1, 3 that one of those would have to be divisible by something?

Actually I think I have just worked it out. If you take any number, it is either a multiple of 3 or you can add either 2 or 4 to make it a multiple of 3. So it's impossible to have three numbers generated by adding 2 and 4 to a given number and not have at least one of them be divisible by 3. Again, I don't know how to write that up as a proper proof.

Why the heck am I studying biochemistry and not maths again? This stuff is so much fun!

*only*way to do it is {3, 5, 7}.Nevermind, that doesn't work. You could still have a run of primes where the last digits went 7, 9, 1 or 9, 1, 3. I may still have generally been on the right track, though; perhaps there is some way of showing that if the last digits went 7, 9, 1 or 9, 1, 3 that one of those would have to be divisible by something?

Actually I think I have just worked it out. If you take any number, it is either a multiple of 3 or you can add either 2 or 4 to make it a multiple of 3. So it's impossible to have three numbers generated by adding 2 and 4 to a given number and not have at least one of them be divisible by 3. Again, I don't know how to write that up as a proper proof.

Why the heck am I studying biochemistry and not maths again? This stuff is so much fun!

6

reply

Report

#3

You need three consecutive odd numbers, all of which are prime.

As the first one is prime, it is not a multiple of 3. What are the implications of this for the other "primes" in your list? Think about the numbers before and after the first one in your list.

As the first one is prime, it is not a multiple of 3. What are the implications of this for the other "primes" in your list? Think about the numbers before and after the first one in your list.

1

reply

Report

#4

(Original post by

~

**Pangol**)~

0

reply

Report

#5

(Original post by

Bit of a full solution - want to edit it down to a hint?

**DFranklin**)Bit of a full solution - want to edit it down to a hint?

1

reply

(Original post by

This isn't a rigorous mathematical proof, but more a logical narrowing-down. All primes are odd except for 2. No primes end in the number 5, except for 5, because anything else that ends in 5 is divisible by 5 and hence not prime. Any set of primes where each successive prime was 2 more than the last one would have to have last digits from among these, in order: 1, 3, 5, 7, 9. There is no way of doing that without including 5 as a last digit. Hence the

Nevermind, that doesn't work. You could still have a run of primes where the last digits went 7, 9, 1 or 9, 1, 3. I may still have generally been on the right track, though; perhaps there is some way of showing that if the last digits went 7, 9, 1 or 9, 1, 3 that one of those would have to be divisible by something?

Actually I think I have just worked it out. If you take any number, it is either a multiple of 3 or you can add either 2 or 4 to make it a multiple of 3. So it's impossible to have three numbers generated by adding 2 and 4 to a given number and not have at least one of them be divisible by 3. Again, I don't know how to write that up as a proper proof.

Why the heck am I studying biochemistry and not maths again? This stuff is so much fun!

**anosmianAcrimony**)This isn't a rigorous mathematical proof, but more a logical narrowing-down. All primes are odd except for 2. No primes end in the number 5, except for 5, because anything else that ends in 5 is divisible by 5 and hence not prime. Any set of primes where each successive prime was 2 more than the last one would have to have last digits from among these, in order: 1, 3, 5, 7, 9. There is no way of doing that without including 5 as a last digit. Hence the

*only*way to do it is {3, 5, 7}.Nevermind, that doesn't work. You could still have a run of primes where the last digits went 7, 9, 1 or 9, 1, 3. I may still have generally been on the right track, though; perhaps there is some way of showing that if the last digits went 7, 9, 1 or 9, 1, 3 that one of those would have to be divisible by something?

Actually I think I have just worked it out. If you take any number, it is either a multiple of 3 or you can add either 2 or 4 to make it a multiple of 3. So it's impossible to have three numbers generated by adding 2 and 4 to a given number and not have at least one of them be divisible by 3. Again, I don't know how to write that up as a proper proof.

Why the heck am I studying biochemistry and not maths again? This stuff is so much fun!

(Original post by

You need three consecutive odd numbers, all of which are prime.

As the first one is prime, it is not a multiple of 3. What are the implications of this for the other "primes" in your list? Think about the numbers before and after the first one in your list.

**Pangol**)You need three consecutive odd numbers, all of which are prime.

As the first one is prime, it is not a multiple of 3. What are the implications of this for the other "primes" in your list? Think about the numbers before and after the first one in your list.

Is there a concrete way to show that if a prime ends in a 7 or a 9, then either or will not be prime?? I can't seem to show this

0

reply

Report

#7

(Original post by

Is there a concrete way to show that if a prime ends in a 7 or a 9, then either or will not be prime?? I can't seem to show this

**RDKGames**)Is there a concrete way to show that if a prime ends in a 7 or a 9, then either or will not be prime?? I can't seem to show this

0

reply

(Original post by

Doesn't the result we've been talking about tell you that if p is a prime then either p+2 or p+4 will not be prime, regarless of the last digit of p? Have I misunderstood you, or are you interested in a different kind of proof?

**Pangol**)Doesn't the result we've been talking about tell you that if p is a prime then either p+2 or p+4 will not be prime, regarless of the last digit of p? Have I misunderstood you, or are you interested in a different kind of proof?

0

reply

Report

#9

(Original post by

Actually I think I have just worked it out. If you take any number, it is either a multiple of 3 or you can add either 2 or 4 to make it a multiple of 3. So it's impossible to have three numbers generated by adding 2 and 4 to a given number and not have at least one of them be divisible by 3. Again, I don't know how to write that up as a proper proof.

**anosmianAcrimony**)Actually I think I have just worked it out. If you take any number, it is either a multiple of 3 or you can add either 2 or 4 to make it a multiple of 3. So it's impossible to have three numbers generated by adding 2 and 4 to a given number and not have at least one of them be divisible by 3. Again, I don't know how to write that up as a proper proof.

0

reply

I find myself coming back to this but with a different question.

I need to formulate and explain the necessary conditions on for the existence of such sets of primes, for general . I should give conditions on that allow the existence of (a) one such set, and (b) more than one such set, for any given . These conditions must be necessary but may not be sufficient.

Not quite sure where to start with this one.

I need to formulate and explain the necessary conditions on for the existence of such sets of primes, for general . I should give conditions on that allow the existence of (a) one such set, and (b) more than one such set, for any given . These conditions must be necessary but may not be sufficient.

Not quite sure where to start with this one.

0

reply

Report

#11

(Original post by

I find myself coming back to this but with a different question.

I need to formulate and explain the necessary conditions on for the existence of such sets of primes, for general . I should give conditions on that allow the existence of (a) one such set, and (b) more than one such set, for any given . These conditions must be necessary but may not be sufficient.

Not quite sure where to start with this one.

**RDKGames**)I find myself coming back to this but with a different question.

I need to formulate and explain the necessary conditions on for the existence of such sets of primes, for general . I should give conditions on that allow the existence of (a) one such set, and (b) more than one such set, for any given . These conditions must be necessary but may not be sufficient.

Not quite sure where to start with this one.

p, p+d, p+2d, p+3d, p+4d, p+5d.

For many numbers - such as d = 23 - at least one of the above numbers will be a multiple of 5 and so not prime (unless it's 5 itself).

So firstly why is this the case? Secondly what (if anything) was special about 23? Then how can you generalize this?

0

reply

(Original post by

Thinking out loud a little - haven't solved the problem - but consider the case when k = 6 so that we have the numbers

p, p+d, p+2d, p+3d, p+4d, p+5d.

For many numbers - such as d = 23 - at least one of the above numbers will be a multiple of 5 and so not prime (unless it's 5 itself).

So firstly why is this the case? Secondly what (if anything) was special about 23? Then how can you generalize this?

**RichE**)Thinking out loud a little - haven't solved the problem - but consider the case when k = 6 so that we have the numbers

p, p+d, p+2d, p+3d, p+4d, p+5d.

For many numbers - such as d = 23 - at least one of the above numbers will be a multiple of 5 and so not prime (unless it's 5 itself).

So firstly why is this the case? Secondly what (if anything) was special about 23? Then how can you generalize this?

So can be expressed where is an integer between 1 and 4

Now for your case of (which is also a prime itself, but I'm not sure how this fact helps, if at all)

Now obviously one of must be equal to 0 thus a member is a mult of 5.

I'm having trouble with the actual generalisation of this but I do understand it

0

reply

Report

#13

(Original post by

I'm having trouble with the actual generalisation of this but I do understand it

**RDKGames**)I'm having trouble with the actual generalisation of this but I do understand it

PS note your argument above works as well for d=28 and d=33 as for d=23. And the argument wouldn't have been much different for d=22.

0

reply

(Original post by

I suppose one way to start would be to try to discount what d won't work for k=6 (or maybe a specific larger k). Note I didn't have to focus on mod 5, it just served my purposes.

PS note your argument above works as well for d=28 and d=33 as for d=23. And the argument wouldn't have been much different for d=22.

**RichE**)I suppose one way to start would be to try to discount what d won't work for k=6 (or maybe a specific larger k). Note I didn't have to focus on mod 5, it just served my purposes.

PS note your argument above works as well for d=28 and d=33 as for d=23. And the argument wouldn't have been much different for d=22.

0

reply

Report

#15

0

reply

(Original post by

Exactly. But remember I might have chosen to look at things mod 3 to address d=5.

**RichE**)Exactly. But remember I might have chosen to look at things mod 3 to address d=5.

And in general, for there to exist one such set for any given , is a necessary condition??

0

reply

Report

#17

(Original post by

So then couldn't be a multiple of 3.

And in general, for there to exist one such set for any given , is a necessary condition??

**RDKGames**)So then couldn't be a multiple of 3.

And in general, for there to exist one such set for any given , is a necessary condition??

0

reply

(Original post by

Can't we do better than that? Going back to the previous k=6 example can't we exclude d being coprime with 6,5,4,3,2? (At least for there being more than one such set.) This then is equivalent to not having 2 or 3 or 5 as a factor.

**RichE**)Can't we do better than that? Going back to the previous k=6 example can't we exclude d being coprime with 6,5,4,3,2? (At least for there being more than one such set.) This then is equivalent to not having 2 or 3 or 5 as a factor.

0

reply

Report

#19

(Original post by

I don't quite understand this. So cannot have primes up to as factors in order for there to exist more than one such set? Why?

**RDKGames**)I don't quite understand this. So cannot have primes up to as factors in order for there to exist more than one such set? Why?

0

reply

(Original post by

Go back to the k=6 case. If you consider each of p, p+d,...,p+5d mod 5 then you get a full set of residues mod 5 when d is not a multiple of 5. So the list can all be primes only by p being 5 - certainly you can't then have two such lists.

**RichE**)Go back to the k=6 case. If you consider each of p, p+d,...,p+5d mod 5 then you get a full set of residues mod 5 when d is not a multiple of 5. So the list can all be primes only by p being 5 - certainly you can't then have two such lists.

0

reply

X

### Quick Reply

Back

to top

to top