June 9, 192 1] 



NATURE 



465 



the distance of the occulting screen, and obtains 

 the distance 140 parsecs, which is about four times 

 that estimated for the Hyades. Since the angular 

 extent is 30°, this implies a real length of 

 70 parsecs, and gives a vivid idea of its gigantic 

 dimensions. 



Dr. Pannekoek next proceeds to make an esti- 

 mate of the mass of the nebula; he first works on 

 the assumption that it is composed of hydrogen. 

 The star-counts indicate an absorption of two 

 magnitudes. Using Dr. Abbot's absorption co- 

 efficient for atmospheric air, and taking the thick- 

 ness of the stratum as 10 parsecs, he finds 10-" 

 for the density of the gas-cloud. The mass is 

 independent of the thickness assumed, and comes 

 out twenty thousand million times that of the sun. 

 This is greater than many estimates of the com- 

 bined mass of the whole sidereal system, and at 

 once suggests the probability that the larger por- 

 tion of that mass is not condensed into stars, but 

 distributed in cosmic clouds. Some striking con- 

 sequences of the presence of such a great mass at 

 a comparatively moderate distance are drawn by 



Dr. Pannekoek. The sun would travel round it 

 in a very eccentric ellipse in a period of some 

 two million years, being now near apocentron. 



While so large a mass of the universe as a 

 whole would render it easier to account for the 

 numerous cases of high velocity among the stars, 

 it would make the moderate velocities of the bulk 

 of the stars difficult toexplain. The author notes a 

 suggestion by Prof. De Sitter that the necessary 

 mass of the occulting clouds may be greatly re- 

 duced if we postulate that they consist of dust 

 instead of gas. It has already been pointed out 

 that the aspect of the dark marking near 

 ^ Ononis accords with the hypothesis of solid 

 matter. However, even in this case the necessary 

 mass is enormous, since it is a priori improbable 

 that the thickness of the stratum should in all 

 cases be a very small fraction of its visible dimen- 

 sions. 



The demonstration of the existence of these 

 immensely massive cosmic clouds seems to make 

 it desirable to rediscuss the dynamics of the stellar 

 svstem. 



Obituary. 



Prof. E. B. Rosa. 

 'T*HE death, on May 17, of Prof. Edward Bennett 

 -*- Rosa, of the Bureau of Standards at Wash- 

 ington, at the age of sixty years, is a serious blow 

 to electrical science. Born in 1861, Prof. Rosa 

 gained distinction as a student in the Johns Hop- 

 kins University under Rowland, and after some 

 experience in professorial work in the Wesleyan 

 university where his early undergraduate days 

 nad been passed was in 1901 appointed to the 

 staff of the Bureau of Standards as a physicist. 

 There his main work was done. In 1910 he 

 became chief physicist, and as head of the elec- 

 trical department was responsible for many 

 of the valuable researches which have been carried 

 out at the bureau. 



Among the earliest of these was his determina- 

 tion, in collaboration with Dr. Dorey, of "v," 

 the ratio of the electrical units, and most of them 

 turn on questions relating to the measurement of 

 the fundamental units, the ohm, the ampere, and 

 the volt. He combined in a marked degree the 

 insight required to design and carry through to a 

 successful result a difficult experiment and the 

 mathematical skill needed to develop to a high 

 degree of accuracy the theory on which the ex- 

 periment is based. 



Prof. Rosa's papers on the calculation of co- 

 efficients of self- and neutral induction, and on 

 the theory of the instruments employed in absolute 

 measurements, will always be standard ; while his 

 own experimental determinations of some of the 

 fundamental quantities are among the best which 

 have been made. He realised the need for accu- 

 racy and exactness in the processes of measure- 

 ment, whether applied to scientific work or to 

 industry, and he organised the electrical section 

 NO. 2693, VOL. 107] 



of the bureau in a manner which fitted it to re- 

 spond to the requirements both of scientific and 

 industrial research. The list of his papers covers 

 a wide range, and in all of them he added to our 

 knowledge in a substantial manner. 



Prof. Rosa visited England in 1908, acting, 

 along with Dr. Stratton and Prof. Carhart, as 

 one of the American representatives to the Inter- 

 national Electrical Conference, held in London 

 under the presidency of the late Lord Rayleigh. 

 At that conference a formal distinction was drawn 

 between the absolute and the international units 

 of measurement : between the ohm (10'^ absolute 

 C.G.S. units) and the international ohm — the re- 

 sistance at 0° C. of a uniform column of mercury 

 weighing 14-521 grams, and 106300 centimetres 

 in length; or the ampere (10-^ absolute C.G.S. 

 units) and the international ampere — the current 

 which under certain carefully defined conditions 

 deposits from a solution of nitrate of silver a 

 mass of I- 1 1800 milligrams of silver per second. 



Prof. Rosa would probably have preferred to 

 retain as standards for legal purposes the abso- 

 lute magnitudes 10'^ C.G.S. units for resistance 

 and lO"^ C.G.S. units for current, but he 

 accepted the views of the majority of the con- 

 ference, and at a later date lent his valuable 

 assistance in defining accurately the conditions 

 necessary for the realisation of the international 

 ohm, ampere, and volt. As the outcome of the 

 work of the conference, a committee, known as 

 Lord Rayleigh 's Committee, was appointed to 

 define these conditions, and representatives of 

 England, France, and Germany met at Washing- 

 ton and carried out a series of experiments, the 

 results of which have determined the practice of 

 all National S^^^andardising Laboratories. Of the 



