630 



SCIENCE. 



[N. S. Vol. XII. No. 304. 



A piece of ruby glass between the lamp and 

 wires prevents fogging of the plate. 



As designed, the wires of the guiding 

 eye-piece were in the same plane as the 

 photographic plate, but as they were some 

 three inches from the optical axis of the 

 telescope a star's image was a crescent, and 

 therefore unsuitable for purp6ses of accu- 

 rate guiding. Outside this plane the star 

 image, as seen in the guiding eye-piece, 

 changes to an arrow-head whose point is 

 directed to the optical axis of the telescope. 

 As it was found that the focus of the tele- 

 scope changed during long exposures, an 

 image of the guiding star which was sensi- 

 tive to changes of focus was highly de- 

 sirable. Professor Keeler found that be- 

 tween the crescent and the arrow-head 

 there was an image formed by the intersec- 

 tion, at an acute angle, of two well defined 

 caustic curves in the aberration pattern. 

 The intersection of these caustics offers a 

 very satisfactory image on which to guide, 

 and at the same time is very sensitive to 

 changes of focus. The relation of the 

 plane of the photographic plate and of the 

 guiding threads was so altered that when 

 the former was adjusted to accurate focus, 

 by means of a high power positive eye- 

 piece, a star's image assuined this partic- 

 ular form in the guiding eye-piece. By 

 noting carefully at the commencement of 

 the exposure the form of the star's image, 

 the focus could be corrected by means of 

 the focusing screw as changes were seen to 

 occur. Photographs of four hours' dura- 

 tion were secured on which the star disks 

 near the center of the plate were almost 

 perfectly round, the smallest disks being 

 from 2" to 3" in diameter. 



In compensating for the variations of the 

 motion of the telescope from that of the 

 stars by moving the plate-holder, there is 

 a limit which Professor Keeler has pointed 

 out, to the amount which the plate-holder 

 may be moved without causing distortions 



in some of the starimages. This distortion 

 arises from the fact that the motion of 

 the plate-holder is in a straight line, while 

 the stars describe small circles about the 

 pole. Hence compensation by such a 

 method of guiding is exact at the equator 

 only. The amount which the plate-holder 

 may be moved without causing an appre- 

 ciable elongation of the star's image may 

 be found from the formula, 



d = 



cos d, — cos 



in which d is the displacement of the plate- 

 holder ; e the amount of elongation in the 

 star's image which becomes perceptible ; 3^ 

 the declination of the guiding star, and S.^ 

 the declination of the star on the plate 

 farthest from the guiding star in declination. 

 In the Crossley reflector it was found that 

 at a declination of 70° (where many nebulse 

 were to be photographed) the plate-holder 

 could not be moved in right ascension 

 more than 1.0 mm. without causing an 

 elongation of the fainter star images which 

 were farthest from the guiding eye-piece in 

 declination, of an amount equal to their 

 own diameters. There is also a small dis- 

 tortion in declination, but on the scale of 

 the Crossley photographs it is negligible. 



To prevent halation in the long ex- 

 posures, the plates are backed with a coating 

 of Carbutt's ' Columbian Backing,' which 

 has proved very satisfactory. 



One of the earliest photographs obtained 

 by Professor Keeler was a very successful 

 one of the great nebula in Orion. This 

 and similar photographs pointed to the 

 great efficiency of the instrument for show- 

 ing the structure in the nebulse, and led to 

 the systematic photographing of all the 

 brighter ones within reach of the telescope. 

 This program had been about half com- 

 pleted by Professor Keeler before his un- 

 timely death. In the prosecution of this 

 work, great numbers of faint nebulse were 



