310 Sir Howard Grubb [May 25, 



specimen photograph which he sent over. " The picture is sent in 

 corroboration of a fact I have suspected for some time, viz. that for 

 stellar photographs, after a certain period of exposure, it is quite 

 immaterial whether the atmospheric definition is good or bad, the 

 photographic images of stars will be equally sharp in either case. 

 That good measurable pictures can be taken on nights when refined 

 eye observations of any value are impossible is a very remarkable 

 fact, and one that a priori would probably be deemed unlikely. The 

 explanation appears to be that the discs on the developed film which 

 represents stars are very much larger than the minute circle formed 

 by the converging cone of rays from the object glass at its inter- 

 section with the plane of the film. These discs are produced by so- 

 called photographic irradiation; in other words, by chemical action 

 set up in the film, having origin in the central point of light, and 

 extending gradually and symmetrically over a wider radius from 

 that centre. This being so, whenever the radius of the disc becomes 

 greater than the radius of extreme oscillation of the optical image 

 from a mean point, the resultant photographic action produced by 

 the rapidly moving point of light becomes identical with the effect 

 produced by a similar steady point of light occupying the same 

 mean position." 



Again, in the case of some of the spectroscopic methods of 

 observation, more particularly when a slit is used, this peculiarity of 

 the photographic method is perhaps still more apparent. 



On this subject Dr. Gill, in a lecture delivered in this theatre 

 just three years ago, says, " On account of irregularities in atmo- 

 spheric refraction, the image of a star in the telescope is rarely 

 tranquil, sometimes it shines brightly in the centre of the slit, 

 sometimes barely in the slit at all, and the eye becomes puzzled 

 and confused. But the photographic eye is not in the least disturbed ; 

 when the star image is on the slit the plate goes on recording what 

 it sees, and when the star is not on the slit the plate does nothing, 

 and it is of no consequence whatever how rapidly these alternate 

 appearances and disappearances recur. The only difference is that 

 when the star is steady and the star's image therefore always on the 

 slit, the exposure takes less time than when the star is unsteady. 

 That is one reason why the Potsdam results, in the determination 

 of stellar motion in the line of sight, are so accurate. And there 

 are many other reasons besides, into which I cannot now enter. 

 What, however, it is important to note is this, that w r e have here a 

 method which is to a great extent independent of the atmospheric 

 disturbances which in all other departments of astronomical obser- 

 vation have imposed a limit to their precision." 



Those who are familiar with the use of large telescopes, know 

 only too well that the larger the aperture the fewer are the oppor- 

 tunities on which it can be used with advantage, and the question 

 has often been discussed as to whether the useful limit of aperture 

 has not already been reached, except in cases when it may be possible 



