492 



NATURE 



[FEiB^ARY 8, iyi2 



At the same mceiing Mr. J. A. A. Buu^*!) discu»iipd the 

 preparation of invar tapes, and quoted the results of ex- 

 periments mad«> to determine their variation due to ageing, 

 rolling and unrolling, and tension. The ex|>eriments were 

 very instructive, though the number quoted was smull. He 

 referred to the difficulty of determining the true tempera- 

 ture of a tape suspended in air, and advocated its deter- 

 mination by means of its electrical resistance. Mr. 

 H. K. E. Keeling directed attention to this difftculty in 

 10 lo. and showed that there might be as much as 2° C. 

 between the tape and the surrounding air. 



Referring to the latter on concentric joints in iic by 

 Mr. H. J. F. Gourley in Nature of January 25, Julia R. 

 •Grugell, of Burleigh .Stroud, Gloucestershire, sends some 

 observations on the same phenomenon. The roadside pool 

 studied was 6 feet long, 4 feet broad, and i foot deep in 

 the, middle and about 3 inches at the side. The rings seen 

 only followed the line of the edge on one side, and half- 

 way round the ends, projecting stones apparently inter- 

 fering with their formation on the other side. A piece of 

 ice, 16 inches square, cut out and examined showed the 

 bulb formation. The ice w-as 2 inches thick at the edge, 

 and the first bulb was 3^ inches from the edge and pro- 

 jected 2 inches, being followed by four others of varying 

 shapes. Along the line of the rings the bulbs appeared 

 to be continuous, except the second, which was saucer- 

 -shaped in appearance. Looking at the surface of the pool, 

 which was of Very smooth ice, the rings were seen 

 alternately clear and opaque, and inspection of the under- 

 surface of the piece removed showed the bulbs were clear 

 through to the surface. The opaque effect seemed to be 

 due to ice crystals on the under-surface of the ice. The 

 ice was the result of several nights' frost, with no thaw 

 by day. 



In connection with the exceptionally warm weather of 

 last summer in Europe and the United States, an article 

 by Mr. R. C. Mossman in Symons's Meteorological 

 Magazine for January, on the " Abnormal Weather in 

 "South America during 191 1," is of considerable interest. 

 For the region north of lat. 40° S. in Argentina and part 

 of Brazil the following mean temperatures below the 

 normal (1898-1907) are given for the four months June- 

 September : — mean max. 25° F., mean min. 3-0°, mean 

 temperature 2-7°. From a map showing the departures of 

 the mean temperature for the three months June-.^ugust 

 in Argentina it is seen that the greatest depression, viz. 

 5°, was in a small area in the province of Corrientes. 

 South of latitude 35°, except for a small patch in Buenos 

 Aires, the depression did not exceed 2°, and south of 41° 

 the temperature was above the average, reaching an excess 

 of 3-5° near the Atlantic entrance to the Straits of Magellan. 

 The author observes that there is little doubt that during 

 Ihe year in question a marked displacement of the 

 " centres of action " of both hemispheres took place. In 

 South America the meteorology of the whole year pre- 

 sented a sequence of abnormal features, among the most 

 prominent being excessive rainfall at Rio de Janeiro in 

 "March, 1911, and extreme drought in Buenos .Aires in the 

 same month. 



A COURSE of lectures delivered before the students of the 

 Johns Hopkins University by Dr. Charles Edward Brooks, 

 on " Tables of Mortality and the Theory of Probability," 

 is published in abstract in the Johns Hopkins University 

 Circular, No. 10. The abstract of the six lectures onlv 

 occupies forty-two pages, and is accompanied by a reprint 

 of the " American Experience " table of mortality. It 

 sliould enable any student possessing an elementary know- 

 NO. 2206, VOL. 88] 



ledge of algebra and the mere notation uf the calculus i< 

 solve ordinary problems on life insurance with a ver> 

 small amount of preliminary study. It shows that th' 

 study of probabilities and statistics is not nearly s< 

 difluult as it commonly supposed. It is highly desirable 

 that schooHxiys should learn the elementary notions of 

 probability soon after they have learnt to multiply vulgar 

 fractions, and that the elements of statistical mathematics 

 should be taught in schools and colleges in place of much 

 of the present useless algebraical drill. 



A REi"ORT on graduate work in mathematics in the United 

 States universities and colleges is published in the Bulletin 

 of the American Mathematical Society for December, 191 1. 

 and forms part of the report of the International Com- 

 mission on the Teaching of Mathematics. It would 

 appear that forty years ago practically no facilities existed 

 for the study of higher mathematics, the school established 

 at Harvard under Benjamin Peirce being one of a few 

 noteworthy exceptions. -At present a much more satisfac- 

 tory state of affairs has been reached, many institutions, 

 both large and small, possessing well-attended schools of 

 higher study frequented by both graduate and under- 

 graduate students. The report deals, further, on the 

 advantages and disadvantages of study abroad, the former 

 including the acquisition of foreign languages. In regard 

 to the training of teachers in higher mathematics, the 

 report reveals a very different state of affairs from that 

 now prevailing in England, namely, an excess of demand 

 over supply. This, in the opinion of the committee, is 

 likely to have a disastrous effect in encouraging graduates 

 of mediocre ability to undertake some work that they can 

 call *■ research " in order to qualify as college instructors. 



With the aid of the Rumford Fund, Mr. P. W. Bridg- 

 man, of the Jefferson Physical Laboratory^ 'Of Harvard 

 University, has continued his experiments on the proper- 

 ties of substances under pressures up to 12,ock> atmo- 

 spheres. His most recent work relates to mercury, and 

 the results are published as Memoir 12 of vol. xlvii. of 

 the Proceedings of the American Academy. He uses a 

 steel piezometer, through the bottom of which a minute 

 hole is made. The piezometer is filled with mercury, and 

 is subjected to pressure in a cylinder of water. As the 

 pressure is applied, water is forced into the piezometer. 

 When the pressure is withdrawn mercury flows out, and 

 the amount which remains In the instrument allows the 

 compression to be determined if those of the water and 

 the steel are known. That of steel was determined in- 

 dependently by the change of length of a rod under 

 pressure. That of water was measured by inverting the 

 piezometer, filling it with water, and subjecting it to 

 pressure in a cylinder of mercury. As the experiments 

 were carried out at temperatures from —39° C. to 4-22° C, 

 they allow the isothermals for solid and liquid mercury 

 between these limits to be drawn on a pressure-volume 

 diagram. 



Vol. vi. of the Journal of the Institute of Metals con- 

 tains the May lecture of 191 1, delivered by Dr. G. T. 

 Beilby. The title, " The Hard and Soft States in 

 Metals," gives a very inadequate idea of the importance 

 of the conclusions which the author has reached by his 

 investigations on metals and other substances. The central 

 fact appears to be that the molecules of any pure sub- 

 stance may be assembled in two different ways," each of 

 which imparts to the substance distinct physical properties. 

 In the first form the substance is crystalline, stable under 

 heat treatment, but yields readily to mechanical forces. In 

 the second the substance is amorphous, becomes crystal- 



