72 



NATURE 



[March 22, 1917 



unless the melting-pots are arranged for stirring. 

 There are no sand beds at present known in this 

 country from which sand for optical glass making 

 can be obtained of such uniformity and at so low a 

 cost as that from Fontainebleau. There are, how- 

 ever, large supplies of sand suitable for good flint 

 glass, laboratory glass, table ware, plate glass, etc. 

 Dr. Boswell points out that the whittsi and best sand 

 is almost invariably found associated with car- 

 bonaceous matter and often in the coal measures. 



A USEFUL contribution to our knowledge of alternat- 

 ing stress is contained in a paper read at the Institu- 

 tion of Mechanical Engineers on February i6 by Dr. 

 William Mason, of Liverpool University. The 

 machine used in the experiments was designed by 

 the author so as to be capable of applying repeated 

 torsion, or bending, or repeated torsion and bending 

 simultaneously. It is a slow-speed machine, and most 

 of the work was carried out at frequencies either less 

 than, or very little exceeding, 200 cycles per minute. 

 This plan has several advantages to recommend it. 

 Typical curves were obtained showing how the cyclical 

 range of strain varies with the range of stress and 

 with the number of cycles endured. The material 

 used was a dead mild steel. The range of stress under 

 which the elasticity became impaired was always 

 fairly definite. It was also found that a large number 

 of repetitions were endured at ranges of stress that 

 induced considerable ranges of extra-elastic strain. 

 It is probable that the range of stress that would 

 cause fracture in, say, 100 million cycles will be nearer 

 in amount to the range that actually produced frac- 

 ture in I to 5 million cycles than to the range of 

 stress that brought departure from the elastic con- 

 dition. The calculated ranges of stress at which 

 cracking occurred were found to be greater for solid 

 than for hollow test-pieces. The effect of giving rest 

 to a specimen in which an extra-elastic range of 

 strain has developed is to reduce the range of the 

 strain ; the effect appears to be similar to that of 

 hardening after strain, and not to be of the nature 

 of recovery of elasticit}-. No real adjustment of elastic 

 limits (with equal + and — stresses) was observed. 

 There appears to be a marked variation, with fre- 

 quency of repetition of cjcle, of the physical state of 

 mild steel subjected to repetition of a higher range of 

 stress than that consistent with unimpaired elasticity, 

 the mobility being greater with higher frequency. 

 For the quality of mild steel used, the range of in- 

 duced maximum shear stress, at which the elasticity 

 becomes impaired, is sensibly the same in both 

 alternating torsion and alternating bending, thus 

 showing agreement with Guest's criterion of elastic 

 failure. 



Messrs. John Wheldox and Co., 38 Great Queen 

 Street, W.C.2, have just issued a catalogue (New 

 Series, No. 78) of second-hand books dealing with 

 chemistry which should be of service to many readers 

 of Nature. It comprises the library of the late Dr. 

 Hugo Muller, and is particularly rich in German 

 works. Complete sets of the Chemical Gazette, 

 the Chemist, Chemical News (to 1916), Journal, 

 Proceedings, and Annual Reports of the Chemical 

 Society of London (to iqi6), the Technologist, and 

 the Journal of the Royal Society of Arts (to 1913) are 

 offered for sale. 



The following volumes are in preparation for appear- 

 ance in the "Cambridge Psychological Library" (Cam- 

 bridge University Press) : — Psychology, Prof. J. Ward ; 

 The Nervous Svstem, Prof. C. S. Sherrington; The 

 Structure of the Nervous Svstem and the Sense Organs, 

 Prof. G. Elliot Smith ; Prolegomena to Psychology, 



NO. 2473, VOL. 99] 



Prof. G. Dawes Hicks ; Psychology in Relation to 

 Theory of Knowledge, Prof. G. F. Stout; Mental 

 Measurement, Dr. W. Brown ; The Psychology of 

 Mental Differences, C. Burt ; Collective Psychology, 

 W. McDougall ; The Psychology of Personality and 

 Suggestion, Dr. T. W. Mitchell ; and The Psychology 

 of Dreams, T. H. Pear. 



OUR ASTRONOMICAL COLUMN. 



Comets 1915a and 1916b. — The following continued 

 ephemeris of comet 1915a (Mellish), calculated for 

 Greenwich mean midnight by Mile. J. Vinter-Hansen,. 

 has been circulated by Prof. Stromgren : — 



Dec). Log ;- Log a 



0-8^02 



0-8597 



0-87I0' I 



1917 R.A. 



h. m. s. o / 



Mar. 22 4 59 26 -f4i 250 

 26 5 o 43 221 

 30 2 10 19-5 08543 



Apr. 3 3 44 ^75 



7 5 5 26 +41 15-9 0-8584 o-88i6 



During the above period the path of the comet is 

 in Auriga.^ 



Prof. Stromgren also reports an observation of 

 comet 1916b (Wolf), made at Copenhagen with the 

 14-in. equatorial on March i. The comet is described 

 as being very faint, and the corrections to the 

 ephemeris were —56s., -^-3'. 



The following continued ephemeris is given by Dr. 

 Crommelin in the February number of the Observa- 

 tory :— 



1917 R.A. Decl. Log r Log .^ Hright- 



h. m. s. , , ness. 



Mar. 22 18 56 26 +2 8 0-3063 0-2989 28 



26 19 5 59 32 



30 15 36 3 59 0-2946 0-2751 33 



Apr. 3 25 16 4 58 



7 35 I 60 0-2833 0-2517 39 



II 44 50 7 3 



15 19 54 43 89 02725 0-2291 45 



19 20 4 40 9 16 



23 14 41 10 24 02622 0-2074 53 



27 20 24 46 -t- II 34 



The brightness on April 3, 19 16, is taken as unity, 

 and no allowance is made for physical causes of 

 brightening. It is expected that the comet will be- 

 come bright enough to be visible to the naked eye 

 during June and July. 



The Sun-spot Zones. — The results of an examina- 

 tion of the latitudes of sun-spots, as recorded in the 

 Greenwich observations for the period 1879 to 191 1, 

 have been given by H. .\rctowski {Mem. Soc. Spett. 

 Ital., February, 1917). As in the previous discussions 

 by Mr. Maunder and Dr. Lockyer, he finds that the 

 curve showing the variation of latitude during an 

 ii-year cycle does not fall continuously from high to 

 low latitude, but has several subsidiary maxima and 

 minima. The phase of the cycle is shown to be 

 different in different zones. Thus, in the period 

 1889-1903, the maximum frequency of spots in the 

 zones 20° to 30° occurred in 1892 ; in the zones 20° to 

 10° it occurred in 1893 ; and in the zone + 10° to 

 — 10° in 1894. Spots ceased to be visible in the first 

 of these zones in 1896, in the second they remained 

 until 1901, while in the third they did not appear until 

 1891 and persisted to 1903, two years after the new 

 cycle had begun. The curves suggest that there is 

 a superposition of a number of distinct variations, of 

 which the principal corresponds to the simplified 

 Sporer's curve and extends throughout the cycle, 

 while some persist onlv a few years, and others may 

 be of very short duration. The variation would thus 

 appear to proceed by a succession of impulses. 



