388 PROCEEDINGS OP THE AMERICAN ACADEMY 



" For the sake of comparison, we will quote from the Posiliones 

 Medice of Professor Struve the following measurements of the same 



stars. 



The mean of Struve's Distances is = 14".40 



" " " Positions is := 147°.4 



" " Photographic Distances is = 14''.49 



" " " Positions is = 147°.8 



"The probable error of a single photographic distance is ± 0".12, 

 or quite as small as that attributed by Struve to a single direct meas- 

 urement. The former method has thus in its first efforts attained the 

 limit of accuracy beyond which it is not to be expected that the latter 

 can ever be sensibly advanced. But the photographic process holds 

 out a much better promise. 



" The two principal sources of error by which it is affected are 

 spots on the glass plate, or impurities in the coating in the neighbor- 

 hood of the images, and slight departures from symmetry in their form, 

 as yet noticed only when the plate has been exposed too long to the 

 action of the light. The latter has been the case with most of the plates 

 from which the above measurements have been taken, and they may 

 in consequence be slightly affected. It is certainly to be anticipated, 

 that, by the exercise of more care in regulating the time of exposure, 

 the symmetry of the images can be secured. A microscopic exami- 

 nation will in most cases serve to distinguish accidental spots in the 

 coating, or on the glass, from the molecules, which, by their aggrega- 

 tion, show the action of light. 



" The real difficulty, perhaps insurmountable, which now prevents 

 a most extensive application of photography to astronomy, is the defi- 

 cient sensitiveness of the pi'ocesses in use. Unless photographs of 

 stars as low, at least, as the eighth magnitude can be obtained, its use 

 must be restricted to comparatively few double stars. Should, how- 

 ever, this impediment be overcome, and photographic impressions be 

 obtained from stars between the sixth and eleventh magnitudes, as has 



