Mav 5, 1 910] 



NATURE 



291 



interesting speculations as to the density of the comet, and 

 suggests that golf balls, sown at the rate of two or three 

 per cubic mile, would probably represent fairly well the 

 degree of rarefaction obtaining in the head at the time its 

 cross-section was 12,500 miles. 



In No. 4404 of the Astronomische Nachrichten Herr J. 

 Holetschek discusses the length of the comet's tail at 

 different apparitions, apropos of the question whether the 

 tail, on .\iay 19, will extend far enough to envelop the 

 eaith. In 1759, when the earth passed through the plane 

 of the comet's orbit on May 14, the tail exhibited large 

 fluctuations in apparent length ; on May 5 it was recorded 

 as nearly 47" long, but on May 14 the recorded length 

 was but 19°. Herr Holetschek gives the following values 

 for the length at various apparitions, the first being a 

 mean value, the second the largest value recorded ; the 

 unit is the earth's distance from the sun : — 1456, 0-20, 

 039; 1531, 014, 0-17; 1607, 006, 012; 1682, 010, 0-22; 

 17591 008, ??; 1835, 008, 0-17; in each case the first value 

 is probably the length of the most brilliant, easily seen, 

 part of the tail. For the tail to reach to the earth on 

 May 20 its length must be 0-15 on this scale. Some amount 

 of discussion has appeared in the daily Press as to the 

 probability of its attaining the requisite length, and, to a 

 representative of the Daily Mail, Mr. Crommelin suggested 

 that the chances are about even ; but there is no method of 

 determining the probable length at any particular time, for 

 comets' tails are so very uncertain in their behaviour. 



The Velocity of the Solar System in Space. — The 

 results of a new determination of the velocity of the solar 

 system in space are published by Prof. Stroobant as an 

 extract from the Bulletin de I'Acadetnie roy. de Belgique 

 (No. I, pp. 39-51, 1910). After discussing previous solu- 

 tions of the problem, he takes Newcomb's later value for 

 the apex (A = 277-5°= i8h. 30m.; + 35°), and, from the 

 more recent determinations of radial velocities of stars, 

 calculates the displacement of the sun in that direction. 



From the discussion of the velocities of forty-nine stars 

 situated near the assumed apex, Prof. Stroobant derives 

 1875 km. as the velocity of translation of our system, 

 and from fifteen stars surrounding the anti-apex he derives 

 2155 km. per sec. Combining these results, he finds that, 

 in regard to stars visible to the naked eye, the solar 

 system is travelling towards the assumed apex with a 

 velocity of 1940 km. per second. This value is a little 

 less than that (1989 ±1-52 km.) derived by Campbell 

 taking the apex obtained by himself, and is much greater 

 than that (16-7 km.) found by Kapteyn ; it represents an 

 annual displacement of 4-10 astronomical units. Prof. 

 Stroobant tabulates the stars discussed by him, giving 

 their positions, magnitudes, spectral types, &c., and shows 

 that stars of different types give different values for the 

 velocity of the solar system ; thus twenty stars of the 

 Orion type give a mean value of 225 km., and appear to 

 constitute an individual system in the stellar universe. 



Star Colours. — In a paper which appears in No. 3, 

 vol. xxxi., of the Astrophysical Journal (April, p. 234), 

 Prof. Louis Bell discusses the reputed colours of the 

 comes in double stars in relation to the known facts con- 

 cerning the colours of stars and their spectra. 



It is a fact, established by many investigations, that 

 among the reported colours of double stars there are 

 bizarre tints which are not met with among isolated stars. 

 That this is not due to any physical connection between 

 the comes is shown by the fact that it appears as strongly 

 in the case of optical doubles as in the case of binarv' 

 systems. The suggestion that these tints are merely sub- 

 jective effects of contrast is generally countered by the 

 statement that they are not ahvays complementary, but 

 Prof. Bell shows that this statement is not conclusive. 



From a discussion of the spectra of a number of doubles, 

 and from a number of experiments on artificial stars, he 

 shows, fairly conclusively, that the reported tints are pro- 

 duced physiologically, and have no determined objective 

 existence. As an example of the evidence deduced from 

 the study of spectral type, he mentions 59 Serpentis, where 

 the primary type i. is yellow and the secondary type ii. 

 is blue; a type ii. star of a bright blue colour is unknown 

 among isolated stars, and logically improbable. His 

 experiments show that with artificial stars of unequal 

 magnitudes, such as are found among double stars, , 



NO. 2 II 4, VOL. 83I 



" dazzle tints " and " fatigue " effects probably account 

 for the curious associations of colours met with in the 

 records of the colours of multiple stars and star-clusters. 



The F0R.MAT10N OF Satur.n's Ring System. — In No. 

 4403 of the Astronomische Nachrichten Prof. Lowell dis- 

 cusses the causes which have produced the present con- 

 formation of Saturn's system of rings. He points out that 

 commensurability of period between perturbing and per- 

 turbed masses is a greater factor in determining their k>ci 

 than is the more direct effect of attraction, and shows that 

 in the case of Saturn's system the rings have their present 

 conformations in accordance with this principle. Thus 

 Prof. Lowell shows that, despite its smaller mass, Mimas 

 has been the chief fashioner of the rings, aided by 

 Enceladus and Tethys, in this order. Not only do the 

 older divisions of the rings show this commensurability of 

 period with the satellites, but newly discovered divisions- 

 occur at such distances as would give commensurability 

 the greatest effect. 



THE WATER PROBLEM. 



'T'HE discussion on the constitution of water which took 

 place under the auspices of the Faraday Society on 

 April 26 was remarkable for the presence of two dis- 

 tinguished foreign visitors. Prof. Walden, of Riga, and 

 Prof. Guye, of Geneva, the former having travelled specially 

 from Riga in order to be present, whilst the latter was 

 able to arrange a necessary visit to London in such a way 

 as to enable him to present his paper in person. Contri- 

 butions to the discussion were also received from Mr. 

 Sutherland, of Melbourne, and from Prof. Nernst. 



Prof. Walker, of Edinburgh, occupied the chair, and in 

 opening the discussion remarked on the extreme complexity 

 of the problem of ascertaining the real nature of this 

 commonest of all solvents, and on the great progress that 

 had been made in recent years in the accumulation of 

 quantitative data for its solution. 



Prof. Walden 's paper, "Is Water an Electrolyte?" in- 

 cluded a number of observations that had l>een made in 

 order to determine w-hether water, when dissolved in a 

 medium possessing pow'erful ionising prof)erties, might not 

 itself become an electrolyte. To secure an adequate answer 

 to this question, it was considered necessary to make use 

 (amongst others) of media of which the specific inductive 

 capacity was greater than that of water. The liquids 

 selected were hydrogen cyanide, HCN ; formamide, 

 H.CO.NH, ; nitrosodimethylamine, (CH,),N.NO ; formic 

 acid, H.CO.OH ; and sulphuric acid, the first two solvents 

 being characterised by a specific inductive capacity greater 

 than 84, the value for water. In passing, it may be noted 

 that Prof. Walden 's discovery of the use of formamide as 

 a solvent represents a " find " of extraordinary importance 

 which, even if it stood alone, would form an adequate re- 

 compense for the labour involved in his masterly survey of 

 the wide field of organic liquids ; this solvent appears to 

 reproduce nearly all the valuable qualities of water, in- 

 cluding its convenient freezing point and boiling point and 

 its powerful ionising properties, but will mix freely with 

 important groups of compounds which do not dissolve ti> 

 any marked extent in water. 



Hydrogen cyanide, which with its high specific inductive 

 capacity and great fluidity provides ideal conditions for 

 electrolysis, was found when used as a solvent for water 

 to give molecular conductivities of the order of oooq as 

 contrasted with 300 for a salt such as potassium iodide ; 

 the low molecular conductivity of the water dissolved in 

 hydrogen cyanide finds a parallel, however, in the low^ 

 conductivity of hydrogen cyanide dissolved in water. 

 P'ormamide, with a similar specific inductive capacity, but 

 much smaller fluidity-, gave for the molecular conductivity 

 of water values (about 0016) slightly higher than in 

 the case of hydrogen cyanide. In the case of formic and 

 sulphuric acids the specific inductive capacity is lower than 

 that of water, sulphuric acid being further handicapped by 

 its extraordinary viscosity, but both solvents gave increased 

 values for the molecular conductivity of water, namely, 

 about 0-17 in formic acid, but rising in the case of sulphuric 

 acid as high as 74. From these observations it is clear 

 that the conductivity attributed to the water does not 



