September 12, 1901J 



NA TURE 



469 



system, at any period of the eclipse, can be ascertained from the 

 simplest geometrical considerations. In the case of two 

 spheroids, the computation is not so simple. 



Still a relation does exist between the amount of distortion 

 due to the mutual attractions of two adjacent bodies, and the 

 rate of obscuration in any eclipse, and this relation is capable of 

 discernment and computation. The difficulty, however, does not 

 lie in the computation : it lies in our inability to determine 

 observations refined enough to respond to a demand so exacting 

 as that which necessitates observations correct to within two- 

 hundredths of a magnitude. 



That this degree of accuracy in photometric measurement 

 has been attained to by more than one observer brings the 

 problem of the determination of the figure of a rotating binary 

 system within a reasonable expectation of solution. 



Of the twenty-two Algol variables at present known, five are 

 binary systems the component stars of which revolve in con- 

 tact. It is, therefore, evident that any investigation having as 

 its purpose the figure of the component members of a close 

 binary system should deal first with these five stars. 



Particulars of these stars are as follows : — 



It may be objected that all along it has been assumed that 

 Algol variables are binary systems. What evidence is there that 

 this is so ? 



In theonly cases where independent confirmation is possible — 

 that is, in cases where the stars are bright enough to be dealt 

 with spectroscopically — this confirmation is forthcoming. 



With regard to the foregoing five stars, the spectroscope 

 reveals V Puppis and $ Lyrre to be binary systems. The other 

 stars are beyond the reach of spectroscopic examination, at 

 least with its present limitations. 



Apart from this, however, revolution and consequent eclipse 

 is the only complete explanation of Algol variation. 



In Fig. I is given the light-curve of V Puppis, the first star in 

 the list, and this light-curve may be taken as typical of those 

 of the other four stars. Indeed, the light-curve of U Pegasi 

 (Harvard Circular, No. 23) is practically identical with that of 

 V Puppis. 



The figure of the B Lyrie system has been considered most 

 fully and conclusively by Mr. Myers in the Astrophysical 

 Journal (sa\. vii. p. I); one of his definite results being that 

 each star forming the s)stem is not a sphere but an ellipsoid of 

 revolution. The amount of flattening is found to be 02 in both 

 stars. 



Mr. Myers also deals with the variation of U Pegasi in the 

 same manner (Astrophysical Journal, vol. viii. p. 163), and 

 finds that there is distinct evidence, in the form of the light-curve 

 of this star, of an ellipsoidal figure in both components. 



In the Astrophysical Journal (\o\. xiii. p. 177), the writer 

 considered the variation of V Puppis, the first of the five stars. 

 Again it was found that while, to no insufficient extent, the view 



NO. 1663, VOL. 64] 



that the light-changes were produced by the eclipse of two 

 spheres would meet the facts of variation, an assumption that 

 both components were ellipsoidal in figure would satisfy the 

 observations more fully. 



Since the foregoing article was published, an examination of 

 all the observations of X Carince and RR Centauri made at 

 Lovedale has been completed. It is found that the twin stars of 

 X Carina; have parted company. They are no longer in actual 

 contact, although a distance of only one-tenth of their diameters 

 separates them. The observations of this star also do not indi- 

 cate an unmistakable distortion of either component. 



In the case of RR Centauri we have a twin system similar to 

 that of V Puppis, with this difference, that the form of the light- 

 curve indicates beyond doubt a considerable ellipsoidal form of 

 both stars. Indeed, a dumb-bell figure of equilibrium similar to 

 that indicated in Prof. Darwin's treatise on " Figures of Equili- 

 brium of Rotating Masses of Fluid " (p. 429) would produce 

 variation of the same character as that of RR Centauri. 



Of the interest which attaches to all investigations, whether 

 by telescope or spectroscope, concerning these remarkable 

 binary systems, there can be no manner of doubt. For we are 

 dealing with the origin of stellar systems. 



Hitherto, in theory only have we had cognisance of some 

 great gaseous orb aggregating itself into two elongated spheroids, 

 dividing after the lapse of long ages into two separate and dis- 

 tinct bodies. 



By the action and interaction of their tidal forces, the gap 

 between the component stars grows wider and wider : the systern 

 ceases to be a close binary star fulfilling its period in days ; it 

 takes months to complete its circuit. 



And still the apocentric revolution goes on, until, at last, the 

 star becomes a visual binary, one component separated from its 

 fellow by the width of the whole solar system. 



From V Puppis, on the one hand, a dumb-bell system speeding 

 round in thirty-five hours, to the twin stars of Castor, completing 

 their great round in one thousand years, we have a regular chain 

 of sequences in distance. 



The links of this chain are made evident by observation as 

 well as by theory. It is not unreasonable, therefore, that the 

 present trend of astrophysical research should be in the direction 

 of discovering more fully and certainly the different stages of 

 evolution and development in the architecture of the heavens. 

 Alex. W. Roberts. 



Lovedale, South Africa, August 9. 



A Plea for a Prehistoric Survey of Southern India. 



According to Mr. R. Bruce Foote, and no one is more 

 competent to speak than he, the urgency for the establishment 

 of a genuine prehistoric survey in Southern India is very great, 

 if the study of this most fascinating branch of archaeology is to 

 be encouraged and the wanton destruction of prehistoric monu- 

 ments checked. Such survey, if honestly carried out, would go 

 far to procure much larger data than yet exist as to the dis- 

 tribution over the southernmost districts of the peninsula of the 

 Palaeolithic people whose remains in the shape of chipped stone 

 implements have been found in .so many localities in the 

 Carnatic and Deccan plateau, embedded in Pleistocene deposits. 

 Such data might help materially also to bridge over the great 

 hiatus in time which now appears to exist between the eraof 

 those very rude people and that of the Neolithic tribes which 

 followed them in the same country. 



Further research in the southern districts especially might 

 result in the finding of evidence as to the quarter from which 

 the Dravidian tribes entered the Peninsula — a question of very 

 high ethnological interest. 



Another very important ethnological question might possibly 

 be also answered by such investigations, namely, were the first 

 Dravidian immigrants that settled in Southern India in a 

 Neolithic stage of culture, or must the polished-stone people be 

 considered as pre-Dravidian ? If the question be answered in 

 the latter way, a fresh immigration must be postulated, by 

 which the true Dravidians reached their present country. If 

 the answer affirms the former proposition, the idea of a further 

 immigration may be dispensed with, for the early iron people 

 appear to be the direct descendants of the Neolithic tribes and 

 the ancestors of the present inhabitants. 



Every year numbers of prehistoric burial places are destroyed 

 by the rapacity of the " Waddars," the wandering tribe of tank 

 diggers, who are allowed to annex the fine slabs composing the 



