Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Law of large numbers wikipedia , lookup
Georg Cantor's first set theory article wikipedia , lookup
Large numbers wikipedia , lookup
History of logarithms wikipedia , lookup
Factorization wikipedia , lookup
List of prime numbers wikipedia , lookup
Proofs of Fermat's little theorem wikipedia , lookup
207 TECHNICAL NOTES AND SHORT PAPERS 1. H. Lamb, "On the propagation of tremors over the surface of an elastic solid," Roy. Soc. London, Phil. Trans., ser. A, v. 203, 1904. 2. C. L. Pekeris, "The seismic surface pulse," Nat. Acad. Sei., Proc, v. 41, 1955, p. 469-480. 3. NBS Applied Mathematics Series, 37, Tables of Functions and of Zeros of Functions, U. S. Gov. Printing Office, Washington, D. C, 1954. {MTAC, v. 10, 1956, Rev. 104, p. 249-250.] 4. P. Davis & P. Rabinowitz, "Abscissas and weights for Gaussian quadratures of high order," National Bureau of Standards, Jn. of Research,v. 56, 1956,p. 35-37. {MTAC, v. 11,1957, Rev. 84.] Mersenne Numbers By Hans Riesel During 1957 the author of this note had the opportunity of running the Swedish electronic digital computer BESK in order to examine Mersenne numbers. The intention of the author's investigation on the BESK was to check some known results, and to examine some Mersenne numbers not previously examined. Mersenne numbers are numbers Mp = 2P — 1, where p is a prime. See QlJ, which contains a more complete list of references. The Mersenne numbers have attained interest in connection with digital computers because there is a simple test to decide whether they are prime or composite. This is Lucas's test [TJ. Furthermore the number 2P~1MP is a perfect number, if Mp is a prime, and all known perfect numbers are of this form. In the beginning of 1957 a program for testing the primeness of the Mersenne numbers on the BESK was worked out by the author. This program, using Lucas's test, works for all p < 10000. As a test, this program was run for the following values of p: 2, 3, 5, 7, 13, 17, 19, 31, 61, 89, 107, 127, 521, 607, 1279, 2203, and 2281. The result was that the numbers Mp are prime for these values of p, thus confirming known results. After these tests, values of p > 2300 were to be tested. These values had not been tested before. But since the testing of Mp for one value of p of this order takes several hours on the BESK, a special program for calculating the smallest factor of Mp, if this factor is <10 -220 = 104 85760, was worked out. This special program is based on the following well-known theorem : All prime factors q of MP (p > 2) are of the form q = 2kp + 1, and of one of the two forms q = SI ± 1. The proof of the theorem is quite simple. If g is a factor of Mp, 2P = 1 (mod g). Since p is a prime, since all numbers n for which 2n = 1 (mod g) form a module, and since 2P = 1 (modg), this module consists of all integral multiples of the prime p. Now 2«-1 = 1 (mod q) if g is a prime, and hence g — 1 is a multiple of p, and in fact an even multiple (since g must be an odd number). This is the first part of the theorem: immediately the solution g — 1 = 2kp (k = 1,2,3, •••)• The second part follows from the theory of quadratic residues. Since x2 = 2 (mod g) has x = 2(p+1)/2(modg), we see that 2 is a quadratic residue modg, hence g = ± 1 (mod 8). By of Mp residue found, the above mentioned special program for small factors of Mp (mod g) for all primes g of the theorem were now calculated. was =0, the factor g was printed out. When no factor g < the BESK turned to the next value of p. This program was Received 8 November 1957. License or copyright restrictions may apply to redistribution; see http://www.ams.org/journal-terms-of-use the values When this 10 -220 was run for all 208 TECHNICAL NOTES AND SHORT PAPERS values of p < 10000. The running time on the BESK for one value of p lay between 0 and 2 minutes, depending on the size of the smallest factor. These smallest factors are given below in a table. The known Mersenne primes are also included in the table. Finally, Cole's factor of AÍ67, and Robinson's factors of Afio» and Af167,though greater than 10-220, have been printed in the table. On checking the values against other sources, a misprint in Archibald's note £2] was detected. This misprint, which concerns the value of the smallest factor of Afi63, seems to have come from Kraitchik [3], p. 24 and p. 92. As a further check, all factors in our table have been looked up in Lehmer's Prime Tables, except a few, which are too large, and a separate calculation was made, to check that they are of the form 2kp + 1. Since, however, there are disturbances in digital computers, it is not absolutely sure that all these numbers really are factors of the corresponding Mersenne numbers Afp. Those primes p, for which no factor < 10 -220 of Mp was found, are, except the known Mersenne primes, omitted in the table. When this table had been calculated, the BESK examined the omitted values of p with Lucas's test. This was done for all values of p between 2300 and 3300. Since the available running time for testing Mersenne numbers on the BESK was limited, every value of p was tested only once. The final remainder, see rjlj, was printed out in hexadecimal form. On September 8th, 1957, a run indicated that the number Af32i7 is prime. This result was repeated on September 12th. All other numbers tested turned out to be composite ; however, this result could be false, since the running time is very long. The testing of Af32i7 took about 5h 30m on the BESK, for one run. A table of the smallest factor of 2" — 1, p prime follows. Primes 2P — 1 are indicated by a dash. Ormangsgatan 67C Stockholm-Vallingby, Sweden 1. R. M. Robinson, 842-846. Mersenne and Fermât numbers, Amer. Math. Soc, Proc, v. 5, p. 2. R. C. Archibald, "Mersenne's numbers," Scripta Mathematica, v. 3, 1935, p. 112-119. 3. M. Kraitchik, Recherches sur la Theorie des Nombres, Gauthier-Villars, Paris, 1952. Table of Factors of 2" - 1 : f £ 2 — 3 — 5 — 7 — 11 23 13 — 17 19 37 223 43 431 71 2 28479 107 — 157 8521 33201 29 233 53 6361 31 — 41 13367 67 1937 07721 97 11447 151 18121 23 47 47 2351 61 73 439 109 7459 88807 79 2687 113 3391 59 1 79951 83 167 127 163 1 50287 167 23 49023 License or copyright restrictions may apply to redistribution; see http://www.ams.org/journal-terms-of-use 173 7 30753 89 — 131 263 179 359 TECHNICAL 181 43441 233 1399 283 9623 367 12479 449 12 56303 521 593 1 04369 643 31 89281 757 98 15263 839 26849 883 8831 953 3 43081 1021 40841 1091 87281 1153 2 67497 1223 2447 1321 7927 1433 20063 1481 71089 1543 1 01839 1667 13337 NOTES AND SHORT PAPERS Table of of 2" - 191 383 239 479 311 53 44847 383 14 40847 461 2767 547 5471 601 3607 659 1319 761 4567 857 6857 907 11 70031 967 23209 1031 2063 1093 43721 1181 47 42897 1229 36871 1327 27 30967 1439 2879 1489 71473 1559 3119 197 7487 251 503 317 9511 397 1693 10159 2383 463 11113 557 3343 607 683 1367 773 68 64241 859 72 15601 911 1823 977 8 67577 1033 1 96271 1097 9 80719 1187 2 56393 1231 5 31793 1361 8167 1447 57881 1499 2999 1583 3167 1697 12 35417 r _ 'factor 211 15193 263 23671 337 18199 419 839 487 4871 571 5711 617 59233 701 223 18287 277 11 21297 353 9 31921 431 863 491 983 577 3463 619 1 10183 719 1439 827 66161 877 35081 937 28111 1013 6079 1049 33569 1117 53617 1201 57649 1279 7 96337 811 3 26023 863 82 58911 929 13007 1009 34 54817 1039 50 80711 1103 2207 1193 1 21687 1249 97423 1367 1381 10937 8287 1451 1453 2903 8719 1511 1523 3023 25 22089 1607 1609 28927 23 94193 1709 1721 1 75543 3 79399 License or copyright restrictions may apply to redistribution; see http://www.ams.org/journal-terms-of-use 229 15 04073 281 80929 359 719 443 887 499 20959 587 5 54129 641 35897 743 1487 829 72953 881 26431 35 3 104 16 941 7529 1019 2039 1051 75503 1129 33871 1213 27511 1303 44849 1423 99063 1459 14591 1531 88799 1663 16631 1723 17231 TECHNICAL NICAL NOTES ANE AND SHORT PAPERS Table of Factors Factors of 2P -— 1 : , 1741 10 02817 1847 33247 1931 3863 2003 4007 2081 2 66369 2179 2 48407 2297 57 65471 2399 4799 2539 25391 2617 78511 2689 71 58199 2767 6 25343 2903 5807 3001 32 17073 3121 11 23561 3257 97711 3359 6719 3499 35 82977 3539 7079 3623 7247 3779 7559 10 1 21 72 1777 10663 1861 23551 1973 22327 2011 71881 2083 69671 2203 2333 3 12623 2411 19289 2543 5087 2621 15727 2699 5399 2819 5639 2909 1 10543 3011 20 23393 3137 20 01407 3299 6599 3361 5 57927 3511 35111 3541 7 64857 3659 41 05399 3793 60689 t —Continued 1789 1801 1811 39359 28817 3623 1871 1877 1879 14969 15017 6 05039 1987 1993 1997 67559 11959 3 95407 2039 2063 2017 4079 4127 93 38711 2111 2113 2131 3 41983 22 31329 32 73217 2213 2251 2207 4 00679 1 23593 53113 2339 2351 2389 4679 4703 71671 2417 2447 2459 4919 14503 81 63193 2591 2593 2549 15559 43 63889 7 66937 2677 2657 2663 63913 3 64073 37199 2741 2707 2711 82231 1 73249 5 85577 2857 2837 2843 1 42151 36 79817 22697 2953 2963 2939 88591 5927 5879 3041 3023 3037 24329 18223 6047 3217 3181 3163 19087 4 55473 3329 3323 3319 26633 33191 23 12809 3457 3433 3391 20743 2 98409 44 42303 3527 3529 3517 1 05871 63487 5 62721 3571 3581 3557 64279 21487 4 26841 3733 3719 3701 1 11031 43 51231 13 14017 3823 3803 3797 7607 89 68759 91129 License or copyright restrictions may apply to redistribution; see http://www.ams.org/journal-terms-of-use 1823 1 20319 1913 63 01423 1999 18 07097 2069 3 26903 2141 3 89663 2281 2393 33503 2531 7 89673 2609 36527 2687 1 98839 2753 47 95727 2897 17383 2999 6 71777 3119 24953 3221 6 44201 3347 26777 3491 6983 3533 1 90783 3593 21559 3767 30137 3851 7703 TECHNICAL NOTES AND SHORT PAPERS Table of 3853 33 90641 3931 26 18047 4057 97369 4211 8423 4289 34313 4457 16 31263 4597 27583 4723 29 09369 4861 29167 4957 1 48711 5021 40169 5167 65 31089 5231 10463 5303 10607 5431 54311 5519 6 62281 5639 11279 5791 9 26561 5903 11807 6113 25 30783 6199 61991 3863 7727 3967 56 21 8 58 6 1 3 9 3 20 8 6 1 63473 4073 69617 4217 84407 4297 93777 4483 54799 4603 26009 4759 14217 4871 9743 4967 27823 5039 10079 5171 10343 5233 94271 5347 10127 5437 87809 5521 88337 5683 75183 5801 34807 5939 17657 6131 12263 6203 48873 r 211 _ >rsof 2" - ' factor 3907 5 31353 3989 1 91473 4099 73783 4229 3 29863 4373 61223 4513 1 35391 4639 1 94839 4787 1 14889 4903 49031 4993 79889 5051 10103 5179 2 48593 5237 40 22017 5393 32359 5441 36 56353 5531 9 40271 5711 11423 5807 1 39369 6011 2 88529 6143 8 84593 6263 12527 3917 4 07369 4013 1 20391 4129 15 85537 4271 1 8543 4441 26647 4549 1 36471 4657 5 11 45623 4801 28807 4933 29599 5003 2 10007 5101 81 10591 5197 1 31183 5279 10559 5413 53 26393 5483 25 8 88247 5557 5563 33343 5 34049 5717 5743 34303 5 43217 5843 5849 7 12847 41 05999 6029 6067 84407 2 06279 6173 6163 5 91649 37039 6277 6287 37663 50 29601 3911 7823 4001 24007 4127 74287 4243 01833 4391 8783 4517 27103 4649 95073 4793 67103 4919 9839 4999 09959 5087 40697 5189 55671 5261 42089 5399 10799 5449 17439 License or copyright restrictions may apply to redistribution; see http://www.ams.org/journal-terms-of-use 3923 2 19689 4019 8039 4153 91367 4273 25639 4447 71153 4561 72977 4691 5 72303 4813 28879 4943 9887 5011 80177 5107 51071 5227 11 29033 5297 44 49481 5417 4 33361 5501 6 93127 5623 11 35847 5779 9 24641 5861 46889 6101 79 06897 6197 2 97457 6311 36 47759 212 TECHNICAL NOTES AND SHORT PAPERS Table of Factors of 2P — 1 : factor 6317 6 94871 6389 1 91671 6491 12983 6569 66 21553 6719 12 90049 6883 18 85943 6967 10 72919 7057 6 21017 7207 1 72969 7487 1 79689 7591 18 97751 7691 15383 7853 47119 8017 1 28273 8171 8 00759 8243 16487 8377 50263 8461 14 04527 8629 2 58871 8719 28 77271 8803 28 34567 8951 17903 6323 12647 6397 19 44689 6521 48 90751 6607 80 34113 6737 7 41071 6899 13799 6977 41863 7079 14159 7211 14423 7529 1 05407 7639 5 50009 7741 89 64079 7873 11 80951 8059 70 43567 8179 43 18513 8269 7 27673 8387 92 76023 8467 2 03209 8663 17327 8731 25 14529 8807 2 28983 8963 6 63263 -Continued 6343 6353 90 45119 38119 6421 6449 56 89007 51593 6529 6551 7 31249 13103 6637 6673 52 03409 67 66423 6781 6803 1 08497 2 17697 6917 6947 1 80623 55337 6983 6997 13967 21 69071 7103 7129 14207 18 67799 7297 7411 18 68033 10 22719 7537 7561 7 23553 5 89759 7643 7673 15287 1 84153 7759 7789 1 39663 8 72369 7883 7901 15767 47407 8101 8111 7 12889 16223 8219 8221 19 88999 9 20753 8293 8273 1 99033 10 42399 8423 8419 4 88303 3 53767 8501 8513 2 72417 6 80081 8669 8693 2 60791 26 18039 8741 8737 69929 6 81487 8839 8821 4 05767 65 23183 9007 9001 90071 15 12169 License or copyright restrictions may apply to redistribution; see http://www.ams.org/journal-terms-of-use 6367 63671 6473 27 96337 6553 14 54767 6691 3 21169 6841 41047 6949 68 79511 7039 12 52943 7151 14303 7417 1 18673 7573 45439 7681 39 94121 7823 15647 7919 63353 8161 67 08343 8231 2 14007 8317 3 82583 8429 67433 8597 4 81433 8699 4 17553 8761 1 92743 8867 70937 9029 7 22321 6373 38239 6481 6 22177 6563 13127 6703 4 42399 6871 2 88583 6959 55673 7043 14087 7193 54 09137 7459 1 34263 7589 2 88383 7687 4 45847 7841 23 67983 8011 80111 8167 31 36129 8237 10 54337 8353 50119 8431 8 09377 8599 85991 8713 16 03193 8783 2 28359 8929 1 96439 9059 18119 TECHNICAL 9109 49 55297 9221 7 19239 9343 1 49489 9421 85 73111 9511 95111 9791 19583 9883 1 58129 NOTES AND SHORT PAPERS p - -Continued Table of Factors of 2" — 1-' factor 9127 9137 9157 9173 1 46033 10 05071 30 76753 4 95343 9311 9323 9337 9283 59 41121 14 15273 8 95009 26 14361 9371 9391 9397 9403 18743 93911 2 25529 25 76423 9491 9431 9461 9479 18959 6 79033 75689 5 31497 9521 9539 9601 9613 19079 22 85039 3 61799 57679 9811 9829 9833 9851 78809 77 70313 07689 45 62513 9949 9973 80 98487 99191 A Computation of Some Bi-Quadratic 213 9181 13 77151 9341 74729 9419 18839 9497 5 31833 9619 6 15617 9859 11 04209 Class Numbers By Harvey Conn A fascinating chapter in computational number theory began when Lagrange showed that every positive integer is representable as the sum of at most four perfect squares [lj. Clearly three would not suffice in every case, as 7 = 22 + l2 + l2 + l2 would be an exception ; nevertheless, the problem of expressing some positive integer n as the sum of at most three squares soon achieved a very special role. For, Gauss showed that r(w), the number of such representations, is connected in a very simple way with the much studied (intrinsically positive) class number, h, of the field generated by -\¡—n. Specifically for n square-free (and n ^ 1,3 where h = 1), (D rin) = where g = 12 for n = 1, 2, 5, 6 and g = 24 for n = 3 (mod 8). Thus we could conclude the existence of at least one such representation for the indicated n. Gauss and later, Kronecker, reversed the direction of these equations by making large scale tabulations of h from rin), although, unfortunately, no location for Kronecker's alleged tabulation (for odd n up to 10,000) seems to exist in the literature. In tallying the representation n = Xi2 + x22 + x32 it might be noted that one must count each ordered triple (xi, x2, x3) of positive, negative, or zero integers as a separate unit, so that as much as 23-3! = 48 could be contributed to rin) when such a decomposition into squares is expressed as triples. In more recent times, the representation theory was extended to integers in the field k generated by V5, i.e., to the quantities ß = (a + ¿>V5)/2 where a and b are of the same parity. Here we seek to represent, necessarily, only those integers fi which are positive together with their conjugate (i.e., totally positive). Thus, e.g., a > |ôa/5| > 0. The special surd V5 must be used because then, as Götzky Received 8 January 1958. License or copyright restrictions may apply to redistribution; see http://www.ams.org/journal-terms-of-use