A refinement of Cauchy–Schwarz complexity

Introduction 1. Primes in arithmetic progressions 2. The main theorems II.Oscillation theorems for sifting functions 3. The statements 4. Proof of Theorem B2 5. Proof of Proposition 3.1 6. Proof of Proposition 3.2 7. Proof of Theorem C: preliminaries 8. Proof of Theorem C: completion 9. Proof of Theorem Bl 10.Proof of Theorem B3 III.Oscillation theorems for primes 11.Proof of Theorem Al 12. Proof of Theorem A2 13.Proofs of Theorem A3 and Proposition 2.1 I. INTRODUCTION 1. PRIMES IN ARITHMETIC PROGRESSIONSFor x ~ 2 real, q a positive integer, and a an integer coprime to q, let 0 and ~ be defined by (1.1) O(x; q, a) = E logp = qJZq) (1 +~(x; q, a)), p~x p=amodq

In this paper, we investigate the existence of fixed points that are not necessarily unique in the setting of extended b-metric space. We state some examples to illustrate our results.

In 1995, A. Wiles [1,2] announced a proof of Fermat's Last Theorem, which is stated as follows: If \(\pi\) is an odd prime and \(x\), \(y\), \(z\) are relatively prime positive integers, then \(z^\pi\not=x^\pi+y^\pi.\) In this note, a simpler proof of this theorem is offered. It is proved that if \(\pi\) is an odd prime and \(x\), \(y\), \(z\) are positive integers satisfying \(z^\pi=x^\pi+y^\pi,\) then \(x\), \(y,\) and \(z\) are each divisible by \(\pi\). The proof offered by Wiles uses cyclic groups. It is the purpose of this note to offer a proof of this result, with techniques within the grasp of Fermat.

David Rees completed his Cambridge undergraduate studies in mathematics in summer 1939; in his first three months of postgraduate work in autumn 1939, he produced a characterization of completely 0-simple semigroups. War then intervened: he worked until the end of the war at Bletchley Park, the British codebreaking centre in Buckinghamshire, where he was part of a team that broke the Enigma code regularly for some critical months during 1940. After the war, he first worked at Manchester University, but moved to Cambridge University in 1948. In the immediate post-war period, he continued with research into semigroups and non-commutative algebra. His first paper was very influential, and he is considered by semigroup theorists to be one of the founding fathers of their subject. At Cambridge, after attending a seminar by Douglas Northcott, Rees changed the focus of his research to commutative Noetherian rings. During an extraordinarily productive period between 1954 and 1961, he produced a string of far-reaching, foundational and deep ideas and results of lasting significance. Highlights include reductions of ideals, his Valuation Theorem, the theory of grade, the graded rings that are nowadays known as ‘Rees rings’, the Artin–Rees Lemma and his characterization of local rings whose completions have zero nilradical. Rees was appointed to the Chair of Pure Mathematics at the University of Exeter in 1958 and elected FRS in 1968. He was awarded the Polya Prize of the London Mathematical Society, and an Honorary DSc by the University of Exeter, in 1993. David Rees was born and brought up in Abergavenny; he was the fourth of five children of Gertrude (née Powell) and (another) David Rees, a corn merchant. The family lived above David's father's corn shop. There is history of both longevity and mathematical ability in David Rees's father's line: his father died at the age of 88, three of his siblings had 90th birthdays and one of his great-great-grandfathers was the Reverend Thomas Rees (1774–1858), a well-known non-conformist minister, who, according to one obituarist, was considered to be the best mathematician in Wales in 1802. David Rees was educated at King Henry VIII Grammar School in Abergavenny. At the time, the school had an excellent headmaster, Wyndham Newcombe, who was also a very good teacher of mathematics. Rees's early teenage years were affected by ill health, and he was absent from school for several terms. During those periods of illness, he studied at home independently, and his mother, armed with lists from the young David, became one of the best customers of the Abergavenny public library. This diligence stood him in good stead when he was able to return to normal schooling: under the guidance of mathematics master L. F. Porter, he was able to catch up quickly with his mathematics. He did rather well in School Certificate examinations in 1934 and 1936, and was awarded a State Scholarship and admission to Sidney Sussex College, Cambridge, where his studies were supervised by Gordon Welchman. Rees started as a Commoner, but was made an Exhibitioner after one year and, after he had come top in the Preliminary Examination for Part II at the end of his second year, he was made a Scholar. Rees was persuaded to take Parts IIB and III together in 1939, and another candidate, Hermann Bondi, with whom he had a friendly rivalry and who only had to take Part IIB at that time, managed to just beat him into second place. Rees began postgraduate work in September 1939, without a proper supervisor, but inspired by ‘wonderful lectures’ by Philip Hall. In the autumn of 1939, he had a rather successful three months, during which he produced a characterization of completely 0-simple semigroups. Here are the relevant definitions. Let S be a semigroup, with operation written multiplicatively. A (two-sided) ideal of S is a non-empty subset A of S such that a s ∈ A and s a ∈ A for all a ∈ A and s ∈ S . A zero element of S is a (necessarily uniquely determined) element 0 ∈ S such that 0 s = 0 = s 0 for all s ∈ S . The semigroup S with zero is called 0-simple if { 0 } and S are its only ideals and there exist s , t ∈ S such that s t ≠ 0 . The semigroup S is said to be completely 0-simple if it is 0-simple and has a non-zero idempotent element e such that the only idempotent f ∈ S for which e f = f e = f ≠ 0 is e itself. Paper [1] was submitted in early May 1940, and represents a very successful start by Rees to postgraduate research. Given Rees's intensive work at Bletchley Park from December 1939 (see §3), most of the work for [1] must have been completed in Rees's first three months of research. In that paper, Rees does thank ‘Mr. P. Hall, both for his encouragement, while this paper was being written, and his very considerable assistance in preparing the paper for publication’. It should be noted that paper [1] is explicitly mentioned in the citation that accompanied David Rees's election as FRS. The phrases ‘Rees matrix semigroup’ and ‘Rees Theorem’ ensure that his name will live on among the semigroup community. By summer 1939, Gordon Welchman had been appointed to work at Bletchley Park, the British codebreaking centre in Buckinghamshire. In December 1939, Welchman knocked on the door of Rees's college rooms to tell him that he had a job for him to do. Rees naturally wanted to know details, but Welchman refused to elaborate, and only after prompting did he tell Rees to meet him a few days later at Bletchley railway station. Rees did so, and in this way was recruited to a team of codebreakers in Hut 6 at Bletchley Park. Welchman recruited several other young mathematicians he knew from Cambridge, including some he had taught at Sidney Sussex College. Even in later life after the veil of secrecy that covered the war-time exploits of Bletchley Park had been lifted, David Rees did not like to talk about his time there. However, it is now clear that he was part of a team that broke the Enigma code regularly for some critical months during the summer and autumn of 1940. The German operators of the Enigma machines were told which three of the five available rotors and which settings to use each day, but they had to choose the initial positions of the rotors and indicate their choices by means of the first three letters of their initial messages. John Herivel, who had also been recruited to Bletchley Park from Sidney Sussex College by Welchman, predicted in February 1940 that some German operators, when tired or stressed, might use short cuts that could be exploited by the Bletchley Park codebreakers. For three months, this lateral thinking by Herivel produced no result; but in May 1940 some of the German operators began to make the predicted mistakes, and David Rees and his fellow codebreakers were able to use the technique known as the ‘Herivel tip’ to break Enigma ciphers for some critical months from May 1940. Herivel has written an account 〈7〉 of the Herivel tip and related matters, in which he attributes the first successful use of the tip to David Rees: see 〈7, pp. 118–119〉. Interestingly, the same book contains a reproduction of a statement by David Rees about the Herivel tip in which he declared that he did not recollect being the person responsible for the first successful use of it, although he conceded that ‘it is possible that my memory is at fault’; see 〈7, p. 122〉. What is not in doubt is that the first successful use of the Herivel tip resulted in much rejoicing, shouting and standing on chairs. Rees thought very highly of Herivel's idea: he described it as ‘brilliant’ in the above-mentioned statement 〈7, p. 122〉; and he is quoted in 〈7, p. 11〉 as having said, in 2000, that ‘of course, the Herivel tip was one of the seminal discoveries of the Second World War’. Rees told the present author in 2007 that, in his opinion, Herivel did not receive the recognition that he deserved. In late 1941, David Rees was seconded to the Enigma Research Section at Bletchley Park, run by Dillwyn (‘Dilly’) Knox, and where the Abwehr Enigma used by the German Secret Service was broken. The so-called ‘Double Cross Committee’ used captured German agents to persuade Hitler that the D-Day landings would be south of Calais rather than in Normandy. It is said that, without the break into the Abwehr Enigma, British intelligence officers could not have known that the deception was working. David Rees subsequently moved to the ‘Newmanry’, the department at Bletchley Park led, for the second half of the war, by M. H. (Max) Newman, for which the first Colossus computer was constructed to assist with codebreaking. The list of subsequently famous mathematicians whom David Rees encountered during his service at Bletchley Park includes A. O. L. Atkin, I. J (Jack) Good, J. A. (Sandy) Green (who worked at Bletchley Park as a teenager), Peter Hilton, Max Newman, G. B. Preston and Shaun Wylie. Sandy Green and Peter Hilton were later to become coauthors of mathematical papers with David Rees, and Rees's third paper [3] (written after the war) was about a paper by Jack Good. There are now available in print numerous articles detailing aspects of the war-time exploits at Bletchley Park; two recent ones are The Guardian's obituary of Peter Hilton 〈23〉 and the Royal Society's Biographical Memoir about William Tutte 〈31〉. Following the end of the war, David Rees resumed his academic studies, and soon found himself working under Max Newman in a different context: he was appointed in 1945 to an Assistant Lectureship in the Department of Mathematics at Manchester University, and Newman was the head of that department. Rees remained at Manchester until 1948, when he was appointed to a University Lectureship at Cambridge; in 1949 he was appointed to a Fellowship at Downing College. He worked in semigroup theory and non-commutative algebra while at Manchester, and continued with these themes for his first years as a Cambridge don. He was very pleased with his joint paper [8] (with Sandy Green) from this time; in it they considered, for positive integers n and r with r ⩾ 2 , the semigroup S n , r (again written multiplicatively) generated by n elements in which each element x satisfies x r = x , but which is otherwise free, and they showed that the question as to whether S n , r is finite is intimately related to Burnside's Conjecture in group theory. Recall that the latter conjecture for r is the statement that, for all n > 0 , the group B n r generated by n elements in which each element x satisfies x r = e , but which is otherwise free, is finite. A striking result from the Green–Rees paper [8] is that the Burnside conjecture for r is true if and only if S n , r + 1 is finite for all n > 0 . David Rees wrote just five papers on semigroup theory, but their influence on the development of that subject has been very substantial. Interested readers might like to consult the tribute 〈11〉 to Rees in Semigroup Forum, where he is described as ‘one of the pioneers of semigroup theory’, as ‘one of the subject's founding fathers’, and as having ‘laid the foundations for a number of important avenues of future research’. However, as David Rees published about forty papers in commutative algebra, it is appropriate that the majority of this obituary be devoted to his contributions to that field. Another addition to the Mathematics faculty at Cambridge in 1948 was Douglas G. Northcott, who had spent 21 post-war months in Princeton, where he had been greatly stimulated by a seminar with the title ‘Valuation theory’ run by Emil Artin and Claude Chevalley, and by much informal guidance from Artin. Northcott returned to Cambridge having become a dedicated algebraist (his PhD work concerned a theory of integration for functions with values in a Banach space). In Princeton, Northcott had, at Artin's suggestion, studied the famous paper 〈28〉 by Weil, and, as a consequence, began to work in the algebra underlying what some refer to as the ‘pre-Grothendieck’ era of algebraic geometry. Thus Northcott became a commutative algebraist. Back in Cambridge, Northcott organized a very successful working seminar on Weil's book 〈29〉. David Rees was a member of the audience, and he was so inspired by the seminar that he too became a commutative algebraist. (Another aspect of Northcott's seminar that was life-changing for Rees was the presence in the audience of Joan Cushen: David and Joan were married in 1952.) David Rees's transition from semigroup theory was gradual and his first paper in commutative algebra ([9], written jointly with Northcott) only appeared in 1954. That paper is central to the next section. Paper [9], written jointly with Douglas Northcott, is, by a long way, David Rees's most-cited research paper: Mathematical Reviews records more than 200 citations of it. It introduced the notion of reductions of ideals. This concept and the related concept of integral closure have had a major influence on research in commutative algebra in the more than 60 years since they were introduced; indeed, even in the present century, hardly a top-level international conference in commutative algebra passes without there being several mentions of reductions. Let b and a be proper ideals of R. The ideal b is said to be a reduction of a if b ⊆ a and there exists s ∈ N 0 (the set of non-negative integers) such that b a s = a s + 1 . One can view such a b as an approximation to a that nevertheless retains some of the properties of a: for example, a prime ideal p of R is a minimal prime ideal of b if and only if it is a minimal prime ideal of a, and when that is the case, the multiplicity of b corresponding to p is equal to the multiplicity of a corresponding to p. (The multiplicity of a corresponding to its minimal prime ideal p is the multiplicity e ( a R p of the ideal a R p of the R p The for the of reduction to David Rees while he was thinking about so-called ideals in a graded S = n ∈ N 0 S n that is as an algebra S 0 , by elements of S + = n ∈ N S n N the set of positive A = n ∈ N A n be a graded ideal of R generated by elements of Rees noted that A n = S n for all n is, A is if and only if there exists ∈ N 0 such that A ( S + = ( S + + 1 . This to the of the concept of The between reductions and integral can be as Let b ⊆ a be ideals of R. b is a reduction of a if and only if each element of a is on the set J of all ideals of R that have b as a reduction has a b b is the of the of and this ideal b is the set of all elements of R that are on The ideal b is called the integral closure of it has the that the ideals of R that have b as a reduction are those between b and b . that b is if b = b . The ideal b is said to be a minimal reduction of a if b is a reduction of a and there is no reduction of a with b (the of is written under the that R is a local with and so that will be in until also will the ideal of Rees and Northcott the ( a of this to be equal to the of ( a ( a , where ( a the graded ∈ N 0 a a + 1 of that reduction of a at ( a that a reduction of a is a minimal reduction of a if and only if it can be generated by ( a and that each reduction of a contains a minimal reduction of Thus all minimal of all minimal reductions of a have ( a on to that ( a can be as 1 , , t ∈ a are said to be in a ∈ N and f ∈ R 1 , , t (the of R in t is a of such that f ( 1 , , t ∈ a , then all the of f in if b is a reduction of a, ( b b = t and { 1 , , t } is a minimal set for it that b is a minimal reduction of a if and only if 1 , , t are in ( a is equal to the number of elements of a that are in a, and a ( a ( a a . mentioned the in the of the of reduction and integral closure in the 60 years since Rees and Northcott published are very and related have been studied in The can some of the influence that these ideas of Rees and Northcott have had, and to in commutative algebra by the book by and on integral That book is dedicated to and David a of and about of Rees's contributions to commutative algebra. In this in which to of the commutative Noetherian some graded rings used by Rees to good these rings are to as ‘Rees and Rees A between graded rings is an that There is an between R a and R ( a . that the graded R ( a a R ( a is to the graded ( a = ∈ N 0 a a + 1 of The R ( a is also called the up of this has its in the that the of R ( a is the underlying the by up ( R with to The of R a , 1 is and Rees used to very good the that, for an ∈ N 0 , the of the graded ideal R a , 1 of R a , 1 is just a . In other R a , 1 R = a . By means of this Rees was able to some about of an ideal in a Noetherian to the where the ideal is and generated by a In that case, are The of Theorem, on the in Rees his use of the above that R is r ∈ = 1 a , then there exists a ∈ a such that r = a r . first with the where a is the ideal R generated by a r ∈ = 1 a , for each ∈ N , r = s for some s ∈ R . s = s + 1 for all ∈ N , since is a in R. R s 1 ⊆ R s 2 ⊆ ⊆ R s ⊆ , and so there exists ∈ N such that R s = R s + 1 . Thus s + 1 = s b for some b ∈ R , from which see that s = s + 1 = s ( b , with b ∈ R . r = s = ( b s = ( b r . In the case, the Rees S = R a , 1 , and set = 1 , a of that Let r ∈ = 1 a . r ∈ = 1 S , and, by the first of this can r = f r for some f ∈ S . f = = b , where b ∈ a for all = , , . of 0 to see that r = b 1 r , and that b 1 ∈ a . In the same paper Rees also a of what is now known as Artin–Rees That also the Rees Lemma (The Artin–Rees Lemma Lemma that R is Let a , b be two ideals of R. there exists ∈ N such that a n b = a n ( a b for all n ⩾ . S = R a , 1 , the Rees of a, and B = b R , 1 S , an ideal of Thus an element = r of R , 1 to B if and only if r ∈ a b for all = , , . B is a graded ideal of the Noetherian and so has a finite set of { b 1 1 , , b } where b ∈ a b for all = 1 , , . Let = { 1 , , } . a n b = a n ( a b for all n ⩾ . (The might it to that ( a b a n ⊆ ( a + 1 b a n ( + 1 for integers n , with 0 n David Rees the name of the as David had his of the in but he did not it for until May paper appeared in in the very in which Emil Artin at a conference in about his of the same and result; M. was to as to who should receive the and that ‘it is the Artin–Rees There is a of the for which that if N is a of a generated then there exists ∈ N such that a n N = a n ( a N for all n ⩾ . This result means that the on N by the on is the on The might like to consult well as being well to the of of a ideal a of the Rees R a , 1 can be used to the integral of the of a, it that R a , 1 R = a for each ∈ N 0 = 1 However, Rees's Valuation Theorem, which is the subject of the next also about the integral of of In a of papers published during an productive period from to David Rees what he called his ‘Valuation which can be as the integral of the of an ideal a of R in of uniquely rings are nowadays to as Rees rings’, while the are called Rees related to the Rees is the as a be a proper ideal of the R. The of a is the a R N 0 { } for which Lemma and Lemma the of for each r ∈ R , the The name is in recognition of P. of the of the theory of ideals in work had a influence on The of the and the of integral closure can be in to the where the underlying A is not and the of Lemma see for a ∈ N 0 , an ideal of A and a , a ∈ A , one can that, if a ∈ , then ( a ⩾ , while if ( a > , then a ∈ . For the statement of Rees's Valuation Theorem, the concept of of including in the where R is not a For properties of and the the is to a of the of the R R p for some minimal prime ideal p of R and a of the of R p that is non-negative on R p . the values of and in { } that, for an r ∈ R , have ( r = if and only if r ∈ p . Valuation that R is Let a be a proper ideal of R. there exist 1 , , of R the of and positive integers e 1 , , e , such that the Rees come in an that their are that, in a ideal is and the Theorem, the statement of which the concept of is an integral such that for each prime ideal p of of the p is a = p ∈ ( , p = 1 p each non-zero a ∈ to only of the prime ideals of of (The that R is that R is an integral its integral closure R is a This result is to in the where R is local and to M. in the that R not be In the about the can be used to Rees's Valuation Theorem, will be on the where R is a in that it is to see where the Rees come A in the of the Valuation from a Noetherian to a commutative Noetherian R is the that, for an ideal a of R and r ∈ R , have r ∈ a if and only if r + p ∈ a + p p for each minimal prime ideal p of R. a of Lemma is related to the integral of of have noted in the that if r ∈ a for a ∈ N , then a ( r ⩾ the Valuation can be used to the statement in Noetherian R. for ∈ N and r ∈ R , it is the that r ∈ a if and only if a ( r ⩾ . This of the Valuation is in such as to about whether two ideals a , b of R are that is, such that a s = b t for some s , t ∈ N . Rees thought that was his best paper, but another of which he was and in which also was In that, he a that had been by in and was related to The latter can be as if S the of in n a and if is a of the of of S that contains must the S be generated In the if is a generated of a and S is a generated integral whose of contains must the S be generated himself in that if the of is 1 or then S is generated In Rees constructed an that showed that the to is For he used and very an Rees the of a of the ideal a on was by in The concept of grade, to the theory of is also to David 1 , , t in R are said to a if they a proper ideal and ( ( 1 , , 1 = ( 1 , , 1 for all = 1 , t . (The of this when = 1 is as ( 0 1 = 0 , that is, 1 is a on In Rees that, for a proper ideal a of each in a has equal to the such that R ( R a , R ≠ 0 . all in a have the same and Rees this to be the of His him to quickly that

Si q≥2 est un nombre entier, on designe par Sq(n) la somme des chiffres en base q du nombre entier naturel n et par vq(n) sa valuation q-adique. L’objectif de cet article est d’etudier des sommes d’exponentielles de la forme ∑n≤xexp(2iπ(lmSq(n)+km′Sq(n+1)+θn)) afin d’en deduire certaines proprietes statistiques des nombres entiers n pour lesquels Sq(n) et Sq(n+1) appartiennent a des progressions arithmetiques donnees. Ceci permet d’etendre les resultats obtenus par Gelfond en 1968 et ceux obtenus par Mauduit-Sarkozy en 1996.

This is the first part of a two-part paper. Our paper was motivated by two classical papers: A paper of Sir Charles Wheatstone published in 1844 on representing certain powers of an integer as sums of arithmetic progressions and a paper of J. J. Sylvester published in 1882 for determining the number of ways a positive integer can be represented as the sum of a sequence of consecutive integers. There have been many attempts to extend Sylvester Theorem to the number of representations for an integer as the sums of different types of sequences, including sums of certain arithmetic progressions. In this part of the paper, we will make yet one more extension: We will describe a procedure for computing the number of ways a positive integer can be represented as the sums of all possible arithmetic progressions, together with an example to illustrate how this procedure can be carried out. In the process of doing this, we will also give an extension of Wheatstone’s work. In the second part of the paper, we will continue on the problems initiated by Wheatstone by studying certain relationships among the representations for different powers of an integer as sums of arithmetic progressions.

An elementary proof of the first fundamental theorem of vector invariant theory

A short proof of the middle-levels theorem, Discrete Analysis 2018:8, 12 pp. Let $n$ be a positive integer, and define a bipartite graph where one vertex set consists of all subsets of $\{1,2,\dots,2n+1\}$ of size $n$, the other consists of all subsets of size $n+1$, and we join a set $A$ of size $n$ to a set $B$ of size $n+1$ if and only if $A\subset B$. The following innocent looking question was an open problem for a long time: does this graph contain a Hamilton cycle? That is, can we find a sequence of sets $A_1,B_1,\dots,A_N,B_N$ (where $N=\binom{2n+1}n$) such that each set of size $n$ or $n+1$ appears exactly once, and such that $A_1$ is contained in $B_1$, which contains $A_2$, which is contained in $B_2$, and so on all the way up to $B_N$, which contains $A_1$? For very small values of $n$, one can find Hamilton cycles, and one can even imagine that one is finding them in a potentially systematic way, but in the end all patterns seem to evaporate. One difficulty is that it can be shown that in any Hamilton cycle, each of the $2n+1$ possible edge directions must occur the same number of times. This means that if one tries to use a solution to the $n-1$ case to build a Hamilton cycle for the $n$ case, one will have to chop it up into exponentially many pieces. (This is in sharp contrast with the case of the entire $n$-dimensional cube, where one can easily prove that it contains a Hamilton cycle by taking two parallel Hamilton cycles in copies of an $(n-1)$-dimensional cube, removing an edge from the first and the corresponding edge from the second, and adding two linking edges to form a single Hamilton cycle.) Another approach that one might imagine attempting to use is a probabilistic one, perhaps creating most of a Hamilton cycle randomly and then making a few adjustments at the end in order to complete it. But this kind of argument also runs into problems: the constraints are not too severe to start with, but the more of the cycle one constructs, the stronger they become, and without a clever idea one gets stuck long before completing a cycle. In 2014 the problem, by then over 30 years old, was solved by Torsten Mutze [1], one of the authors of this paper, using a long and technical (and constructive) argument. This paper gives a much shorter and simpler argument that avoids nearly all the technicalities of the first argument and thereby makes Mutze's remarkable result truly accessible for the first time. Very roughly, the idea behind the proof is to begin by finding a certain special collection $\mathcal C$ of vertex-disjoint cycles that includes all vertices. If there were only one cycle in the collection, we would be done, but since that is not the case, we have to join them together. This is done by using an auxiliary collection of 6-cycles: if a 6-cycle intersects the union of two cycles $C_1,C_2$ in a particular way, then the symmetric difference of all three cycles is a single cycle that visits the same vertices as $C_1$ and $C_2$. The authors prove various lemmas about the cycles in $\mathcal C$ and the 6-cycles, which they then use to prove that this process can be continued until there is just one cycle left. (The reason 6-cycles are used, rather than 4-cycles as in the case of the complete cube, is brutally simple: it is easy to check that the middle-layers graph does not contain any 4-cycles.) The proof gives rise to an efficient algorithm for actually finding Hamilton cycles in the middle-layers graph. If you are curious to see what the cycles look like, the algorithm can be run online on [this web page](http://page.math.tu-berlin.de/~muetze/cos/middle.html). [1] Torsten Mutze, _Proof of the middle levels conjecture_, Proc. London Math. Soc. *112* (2016), pp. 677-713. Also available at [arXiv:1404.4442](https://arxiv.org/abs/1404.4442)

This paper is a contribution to the project of determining which set existence axioms are needed to prove various theorems of analysis. For more on this project and its history we refer the reader to [1] and [2].We work in a weak subsystem of second order arithmetic. The language of second order arithmetic includes the symbols 0, 1, =, <, +, ·, and ∈, together with number variables x, y, z, … (which are intended to stand for natural numbers), set variables X, Y, Z, … (which are intended to stand for sets of natural numbers), and the usual quantifiers (which can be applied to both kinds of variables) and logical connectives. We write ∀x < t φ and ∃x < t φ as abbreviations for ∀x(x < t → φ) and ∃x{x < t ∧ φ) respectively; these are called bounded quantifiers. A formula is said to be if it has no quantifiers applied to set variables, and all quantifiers applied to number variables are bounded. It is if it has the form ∃xθ and it is if it has the form ∀xθ, where in both cases θ is .The theory RCA0 has as axioms the usual Peano axioms, with the induction scheme restricted to formulas, and in addition the comprehension scheme, which consists of all formulas of the formwhere φ is , ψ is , and X does not occur free in φ(n). (“RCA” stands for “Recursive Comprehension Axiom.” The reason for the name is that the comprehension scheme is only strong enough to prove the existence of recursive sets.) It is known that this theory is strong enough to allow the development of many of the basic properties of the real numbers, but that certain theorems of elementary analysis are not provable in this theory. Most relevant for our purposes is the fact that it is impossible to prove in RCA0 that every continuous function on the closed interval [0, 1] attains maximum and minimum values (see [1]).Since the most common proof of the Mean Value Theorem makes use of this theorem, it might be thought that the Mean Value Theorem would also not be provable in RCA0. However, we show in this paper that the Mean Value Theorem can be proven in RCA0. All theorems stated in this paper are theorems of RCA0, and all of our reasoning will take place in RCA0.

In discussing unique factorization and ideal theory, C. C. MacDuffee [l, p. 122] cites the multiplicative system of positive integers of the form 1+7& as an example where unique factorization fails, since 792 = 22-36 = 8-99, and 8, 22, 36, 99 are all primes in the system. H. Davenport [2, p. 21] uses positive integers of the form l+ik for the same purpose, with the numerical case 693 = 9-77 = 21-33. In this paper we examine all multiplicative systems made up of arithmetic progressions, and decide the question of unique factorization. For a fixed positive integer ra, let M be a multiplicatively closed system of positive integers such that if xGA2 and y=x (mod ra), y>0, then yG.M. It will be assumed that w is the smallest positive integer which can be used to define M. For example the set M of all positive integers congruent to 1, 3, or 5 modulo 6 is also the set congruent to 1 modulo 2, and in this case ra = 2. We divide the integers 1, 2, • • • , w into two classes: the set A, 0(ra) in number, of those relatively prime to ra, and the others in set B, ra—0(ra) in number.

Pillai's conjecture revisited

n= Zx,+ i=1 The principal tool in the proof is the Hardy-Littlewood circle method, incorporating results and techniques of a powerful new iterative method developed by Vaughan and Wooley ([Va3], [Va4], [VW1], [Wol], [VW2]). In section 2, an algorithm developed in [Fo] for optimizing the parameters in mixed power mean value theorems is generalized and analyzed. Section 3 details a more sophisticated method of generating mixed power mean value theorems, by a limited adaptation of the iterative method itself. The form of these estimates offers many advantages over those of section 2, and provides the key to the elimination of the 16th power from (1.1). These mean value theorems are then applied to the proof of Theorem 1 in section 4. The tools developed here are applicable to a wide range of mixed power representation problems, and we briefly illustrate in section 5 the application to the problem of determining the number of terms required to represent all large n, when the lowest power used is a kth power instead of a square. Throughout, n is a large natural number whose representation as a sum of mixed powers is at issue, and E is an arbitrarily small positive real number. Constants implied by the Landau Oand Vinogradov <<-symbols may depend on E or k. For a real number x, write e(x) for e27iX, [x] for the greatest integer not exceeding x and llxll for the distance from x to the nearest integer. Unless otherwise specified, lowercase Latin letters denote natural numbers and Greek letters denote real numbers. Let A(P, R) denote the set of natural numbers not exceeding P with no

Consider all the arithmetic progressions of odd numbers, no term of which is of the form 2 k + p, where k is a positive integer and p is an odd prime. Erdős ever asked whether all these progressions can be obtained from covering congruences. In this paper, we characterize all arithmetic progressions in which there are positive proportion natural numbers that can be expressed in the form 2 k +p, and give a quantitative form of Romanoff’s theorem on arithmetic progressions. As a corollary, we prove that the answer to the above Erdős problem is affirmative.

We introduce a new, elementary method for studying random differences in arithmetic progressions and convergence phenomena along random sequences of integers. We apply our method to obtain significant improvements on two results. The first improvement is the following. Let l be a positive integer and {u 1 ≥ u 2 ≥ · · · } be a decreasing sequence of probabilities satisfying u n · n 1/(l+1)→∞. Let R = R ω be the random sequence obtained by selecting the natural number n with probability u n . Then every set A of natural numbers with positive upper density contains an arithmetic progression a, a+r, a+2r,..., a+lr of length l + 1 with difference r ∈ R ω . The best previous result (by M. Christ and us) was the condition u n ·n 2−l+1 → ∞ with a logarithmic rate. The new bound is better when l ≥ 4. Our second improvement concerns almost everywhere convergence of double ergodic averages. We construct a (random) sequence {r 1 < r 2 < · · · } of positive integers such that r n /n 2−ϵ → ∞ for all ϵ > 0 and, for any measure preserving transformation T of a probability space, the averages $$\frac{1}{N}\sum\limits_{n < N} {{T^n}{F_1}\left( x \right)} {T^{{r_n}}}{F_2}\left( x \right)$$ converge for almost every x. Our best previous result was the growth rate r n /n (1+1/14)−ϵ → ∞ of the sequence {r n }.

On square-free arithmetic progressions in infinite words