376 



SCIENCE. 



[Vol. XV. No. 385 



gravity of the system is a,=3".33. Computing the co-efficients 

 of — for the thirteen years for which I have spectroscopic results, 

 the following observation equations are obtained: — 



Giving the weight unity to each of these, the following normal 

 equations result: — 



1 



13 y„ 



5.3 - = 



1 



100 



- 5.2 Fo + 2.124 _ = - 39.86 

 Hence tt = 0''.34. '^ 



The only direct determinations of the parallax of Sirius that I 

 have at hand are those of Henderson and Peters (0".150), Gylden 

 (0".193), and Gill and Elkin (0".39). The agreement between this 

 last result and the one deduced above by the spectroscopic method 

 is of course purely accidental, the Greenwich measures being too 

 rough to furnish us with a result even passably good. It is to be 

 hoped that measures may be made, with improved apparatus, 

 that will enable us, in the course of time, to apply the spectro- 

 scopic method to a large number of double stars. Dr. Vogel's 

 photographic measures, taken at Potsdam, are far superior to any 

 thing else the writer has seen in this line, unless the Harvard 

 College measures are excepted. They have a very small probable 

 error, and the measures on Venus seem to indicate that his work 

 is not materially affected by constant errors. As a matter of 

 fact, constant errors are of no importance in deducing parallax 

 by the spectroscope, since they affect Vo only, and not tt. 



It is plain that the method set forth in this article is open to 

 many objections. It is beset -with difficulties, but it should not 

 be discarded or lost sight of for that reason. The fact is, that we 

 must either forever give up the effort to determine the distances 

 to the more remote stars, or we must seek them by this method. 

 Very few stars are near enough for us to measure their parallaxes' 

 directly ; but the spectroscopic method is still applicable when 

 there is no indication of parallax to the mici'ometer. The velocity 

 of motion in the line of sight may be measured with equal ease, 

 whether the star is near or remote; and the only limit to the 

 power of the method lies in the increasing closeness of the double 

 star as the distance grows greater, and the corresponding difficulty 

 of determining the orbits. By photographing binary systems at 

 intervals of a few years, and measuring distances and position an- 

 gles on the negative, much better orbits might be oomputed than 

 we have at present. 



While at work on the numerical example in this article, I have 

 endeavored to find out whether others have not worked at this 

 same problem. My attention has been called to several papers 

 that apparently relate to it, though I have been unable to gain 

 access to more than two of them, and then only for a moment, so 

 that I could give them only a hasty examination. The first paper 

 that I find reference to was by C. Dufour, the title being " Utili- 

 sation de I'analyse spectrale pour determiner la distance de cer- 

 taines etoiles doubles " (Lausanne, Bulletin des seances de la So- 

 ciete vaudoise des sciences naturelles, vol. siii., 1874, p. 453). The 

 second paper, by Edward C. Pickering, is entitled " Dimensions 

 of the Fixed Stars, with Especial Reference to Binaries and 

 Variables of the Algol Type," and is in the "Proceedings of the 

 American Academy of Arts and Sciences," 1880, vol. xvi. This 



paper, as its title indicates, relates more particularly to the dimen- 

 sions of stellar systems than to their parallaxes. The third paper 

 was read before the Royal Irish Academy on May 24, 1886, by 

 Arthur A. Rambaut, the title being " On the Possibility of deter- 

 mining the Distance of a Double Star by Measures of the Relative 

 Velocities of the Components in the Line of Sight." The method 

 here proposed is not applicable unless both components are bright 

 enough to be observed spectroscopically. It appears to be identi- 

 cal with what I have called the "first method." Finally, Herr 

 J. Palisa published an article on the subject while the present 

 one was in preparation, entitled " Ueber die Bestimmung der 

 Parallaxe von Doppelstemen " {Astr. Naeh., No. 2,941, Dec. 13, 

 1889). I have not seen this paper, but I understand that in it he 

 refers to a dissertation by one Hans Homann, the title of which 

 is " Beitrage zur Untersuchung der Sternbewegung," which pre- 

 sumably touches upon the same subject. A. D. Risteen. 

 Hartford, Conn., June ly. 



Temperature in Storms. 



In the " Smithsonian Report for 1865," beginning at p. 340, there 

 is a detailed account of a baUoon ascension at Paris on July 27, 1850, 

 during a severe storm of rain with some strong wind gusts. The 

 most interesting point is the severity of the cold encountered, the 

 temperature falling to — 39° C. at an elevation of seven thousand 

 metres. It is stated that at the beginning of the ascent "a 

 deluge of rain was falling," which shows that it must have been 

 made near the centre of precipitation, if not at the exact storm- 

 centre. It would seem that these observations are confirmatory 

 of those noted by Dr. Hann. to which reference is made by Pro- 

 fessor Davis in Science for May 30. M. A. Veedee. 



Lyons, N.T., June 9. 



BOOK-REVIEWS. 



The Criminal. By Havelock Ellis. London, Walter Scott; 

 New York, Scribner & Welford. 12°. $1.25. 



Mr. Ellis, the editor of this promising series of scientific mon- 

 ographs, has contributed to it in this volume a well-planned and 

 ably executed resume of modem criminology. So little of this 

 science has hitherto been accessible in English, that this compila- 

 tion is especially timely. It is an outcome of a pedantic and un- 

 scientific view of crime, that we are obliged to speak and accus- 

 tomed to think of all persons liable to punishment as criminals. 

 There are certain very well marked distinctions between classes 

 of criminals, that should be generally recognized. There is the 

 criminal by passion, the insane criminal, the occasional criminal, 

 the instinctive criminal, the habitual criminal, the professional 

 criminal. While the existence of some of these is' rather the 

 crime of the society that breeds them, others are distinctly dis- 

 eased forms of humanity, which we must study in order to un- 

 derstand and to treat. It is the biological, sociological, and psy- 

 chological study of the criminal classes, so vigorously pursued in 

 Italy and other countries of Europe, that is described in the 

 present volume. On the physical side, the shape of the cranium, 

 the tendency to asymmetry, the peculiarities of face, the details 

 of the ear, nose, and so on, anomalies of the hair and pigmenta- 

 tion, hereditary characteristics, motor inertia and sensory insen- 

 sibility, fondness tor tattooing, and the like, — these characteris- 

 tics have been made the subjects of special monographs, and in 

 many cases class distinctions between the criminal and his more 

 normal fellow-being have been successfully laid down. While it 

 is not yet possible to describe accurately and briefly the results of 

 these methods of study, and still less so to apply them to indi- 

 vidual cases, enough has been done to indicate that all kinds of 

 abnormalities are more common among criminals than among 

 normal people, and to give interesting glimpses into the nature of 

 these differences. These physical differences are connected with 

 and lead up to moral and intellectual differences, and the studies 

 of the two have advanced together. The moral obtuseness, the 

 lack of sympathy, the selfish aod thoughtless satisfaction of sud- 

 den and strong impulse^the keen cleverness in certain limited 

 directions but general stupidity in every thing else, the emotional 



