October 1-1, 1895.] 



SCIENCE. 



475 



the present shore is another difficulty. The 

 prevailing rounded or elliptical shape is 

 not explained. 



That the sea, when this part of the coas- 

 tal plain rose above it, left numerous in- 

 equalities somewhat similar to the ripple- 

 made pittings seen in the sand in the bot- 

 tom of a gutter after a rain has suggested 

 itself to me. If so, these basin-like pittings 

 — separated from each other by sand ridges 

 highest above the general shore slope on 

 their east side — might have formed the ba- 

 sins for these ' bays.' 



Fuller observation and study is needed 

 before anything but a tentative conclusion 

 maybe reached. Any additional observa- 

 tions or suggestions will be gladly wel- 

 comed. 



L. C. Glenn. 



Darlington, S. C. 



A NEW METHOD OF DETEBSIINING THE MO- 

 TION OF STABS IN THE LINE OF SIGHT. 



A METHOD of measuring the motions of 

 stars in the line of light, which does not 

 require the use of an artificial comparison 

 spectrum, and which is therefore adapted 

 to slitless spectroscopes, has been proposed 

 by Professor Orbinsky, of Odessa (A. N. 

 3289). It is of unusual interest because 

 the object-glass spectroscope, which is so 

 advantageous with respect to simplicity of 

 construction and to the brightness of the 

 spectra which it yields, has never yet been 

 successfully applied to this branch of as- 

 tronomical research. 



The principles on which the method de- 

 pends may be briefly described as follows : 

 If a luminous body is moving in the line of 

 sight, the distance between any two lines 

 in its spectrum is not what it would be if 

 the body were at rest, since the two lines 

 are unequally displaced by the mo- 

 tion. In a normal spectrum the dis- 

 placement of the lower line would be 

 somewhat the greater, although the differ- 



ence would scarcely be measureable under 

 ordinary circumstances, but on account of 

 the increasing dispersion of a prism toward 

 the violet the effect in a prismatic spectrum 

 is revei-sed, and the upper line is displaced 

 more than the lower one. The differential 

 displacement of the Hd and Hj3 lines, in an 

 ordinary prismatic spectroscope, is, in fact, 

 somewhat more than half the absolute dis- 

 placement of the H)' line. Hy measuring 

 this apparent change of dispersion the mo- 

 tion of a star can be determined. 



To avoid the errors attending the meas- 

 urement of large distances on a photograph, 

 and other errors which need not be si)e- 

 cially mentioned here, the spectrum of a 

 star whose motion in the line of sight is 

 known is photographed on the same plate, 

 and the apfiarent change of dispersion due 

 to the motion of the first star is deduced 

 from measures referred to corresponding 

 lines in the spectrum of the .second. The 

 stars selected for purposes of comparison 

 would naturally be bright stars with well- 

 defined lines, and their motions could 

 therefore be accurately determined by the 

 usual methods. Only a comparatively small 

 number of such standard stars would be 

 required. 



For slit spectroscopes it would probably 

 be found that Professor Orbinsky 's method 

 is inferior to the usual one, although Pro- 

 fessor Vogel finds that it can be applied to 

 some of the photographs taken with the 

 Potsdam spectograph. It not onlj^ depends 

 upon a differential effect, and thus reduces 

 the amount of the available displacement, 

 but it requires the measurement of lines 

 which are widely separated, and therefore 

 badly defined in consequence of their great 

 distance from the axis of the camera objec- 

 tive. Even if this lens were constructed 

 with a view to giving a large field the defi- 

 nition would be inferior to that in the 

 center of the field of an objective of the 

 usual construction. 



