September 30, 1922] 



NA TURE 



457 



and important role in the maintenance of efficiency, 

 such as lighting, heating, ventilation, the mode of life 

 led by the worker outside his definite hours of labour, 

 his housing, etc. Many of these factors have been 

 partially examined. Thus Leonard Hill has carried 

 out a great deal of valuable work on the influence of 

 the cooling power of the air. Vernon has collected 

 much interesting evidence, which shows that there is 

 a very definite relation between the efficiency, as 

 measured by output, and the temperature of the 

 working place. The output in the hottest weather was 

 about 30 per cent, below that when the weather was 

 coldest. He also observed an apparent connexion 

 between the relative humidity of the air and the 

 efficiency of the worker. The efficiency, as might have 

 been expected, was apparently greatest when the 

 relative humidity was low. Elton has reported on 

 the influence of lighting in silk weaving. He found 



that the output was lowest when artificial light was 

 used. He stated that even when electric light of 

 sufficient intensity was used, the output was about 

 10 per cent, lower than the daylight value. The actual 

 equipment of the factories, the provision of seats of 

 suitable size, height, etc., the design of the machines, 

 and so on, all play their part, as is shown by the many 

 records, particularly from the United States. 



In other words, the real over-all industrial efficiency 

 of the worker cannot be causally related to any single 

 factor. It is not the mere capacity of the individual 

 to perform so many kilogram-metres of work in a given 

 time with the smallest expenditure of energy. The 

 quest of efficiency calls for the closest and most inti- 

 mate co-operation between the scientific investigator, 

 the employer, and the employee, and it can only be 

 satisfactorily attacked when mutual distrust of motives, 

 capacities, and methods is stilled. 



The Total Solar Eclipse of September 21. 



Bv Dr. A. C. D. Crommelin. 



THE failure of the Christmas Island eclipse expedi- 

 tion is a great astronomical disappointment. 

 Messrs. Jones and Melotte have devoted ten months 

 or more to it, and hoped to secure useful photometric 

 results for connecting the northern and southern stellar 

 magnitude scales in addition to the eclipse work. The 

 climate, however, proved unexpectedly unfavourable^, 

 and practically nothing could be done. 



On the other hand, the conditions appear to have 

 been ideal right across Australia, and enthusiastic 

 reports have come from Wollal (West Coast), Cordillo 

 Downs (centre), and Goondiwindi and Stanthorpe 

 (Queensland). The Einstein problem was studied at 

 Wollal by the Lick Observatory party under Prof. 

 Campbell, and that from Toronto under Prof. Chant. 

 Mr. Evershed also finally selected this station in pre- 

 ference to the Maldives, and is believed to have under- 

 taken the same investigation, in addition, doubtless, 

 to spectroscopic work. Prof. Dodwell, the Govern- 

 ment Astronomer at Adelaide, had the use at Cordillo 

 Dow-ns of a tower telescope lent by the Lick Observa- 

 tory for the Einstein problem ; the New South Wales 

 astronomers were in Queensland and did some spectro- 

 scopic work ; they intended also to make Einstein 

 investigations, but the telegrams do not allude to these. 



It is well to point out that the test of the Einstein 

 theory does not depend wholly on the results of this 

 eclipse. The plates secured in the 1919 eclipse at 

 Principe and Sobral settled definitely that at least the 



half-shift was present, while the two cameras with the 

 best definition gave values very close to the Einstein 

 value. Further, the star-field in that eclipse was the 

 best along the whole extent of the ecliptic, the stars 

 in the present eclipse being much fainter. There are, 

 however, two circumstances that should add weight 

 to this eclipse : (1) that some of the observers were 

 pointing directly on the stars, avoiding the use of a 

 ccelostat or other mirror ; (2) that the plan was being 

 tried of photographing another star-field during totality, 

 thus obtaining an independent scale-value for the 

 plates, which gives a much larger coefficient to the 

 Einstein displacement in the equations of condition. 



Probably weeks or months must elapse before the 

 Einstein, results are to hand. 



The corona is said to have had four long streamers, 

 one extending to three solar diameters, which is more 

 than the average, though by no means a record. 



Prof. Chant reports that the shadow bands were 

 photographed. Prof. Kerr Grant, of Adelaide Uni- 

 versity, made measures at Cordillo by the photo- 

 electric cell of the relative brightness of the sun and 

 the corona. The results, with this very sensitive 

 instrument, should be more trustworthy than previous 

 determinations. 



The next two total eclipses (1923, September, and 



1925, January) are visible in the United States ; 



1926, Januarv, in Sumatra, etc., and 1927 in England 

 and Norway. 



Obituary. 



Prof. Alexander Smith. 



ALEXANDER SMITH, emeritus professor of chem- 

 istry in Columbia University, New York, died in 

 Edinl iurgh on September S, aged fifty-seven. Smith was 

 born in Edinburgh, and entered the University there in 

 1882, where he studied mathematics under Chrystal, 

 natural philosophy under Tait, and chemistry under 

 Crum Brown, graduating as B.Sc. in 1886. During 



NO. 2761, VOL. I ioj 



the following three years he attended the University 

 of Munich, working in Baeyer's laboratory, chiefly 

 under the direction of Claisen, and obtained the degree 

 of Ph.D. in 1889. 



After a year spent as assistant in the chemistry 

 department of the University oi Edinburgh, Smith 

 was offered the chair of chemistry and mineralogy 

 in Wabash College, Indiana, a post which he held for 

 four years. In 1894 lie became assistant professor of 



