46 C. H. McLEOD ON NOTES ON" KERORS IN MERIDIAN, Etc. 



of observers and the constant use of the same instrument by each observer, it cannot be 

 entirely' eliminated. The difficulty of course lies in the lack of constancy in the equation. 

 An improvement may possibly lie in the direction of photography, by the aid of which there 

 is some hope that normal observations may in the future be obtained. Some preliminary 

 experiments in the registration of star transits by photography made at the Harvard College 

 observatory and at Washington a few years ago gave great promise, but the method has as 

 yet remained undeveloped. 



The selection of a star list is also a matter which, though it does not belong directly to 

 the subject of observational errors, may, unless carefully considered, resultin a serious reduc- 

 tion in the accuracy of longitude work. The list should be homogeneous. The " Connais- 

 sance des Temps" or the "Berliner Astronomisches Jahrbuch" are probably the best lists, 

 and as between the two the methods of the work will determine the selection. In the 

 French Geographical service it is the practice to make the final adjustment of the star places 

 used in the reductions dependent to some extent upon the observations themselves. 



Although it is probably impracticable for the two or more observers engaged, to use 

 exactly the same stars, they should undoubtedly work from the same list, and if the time 

 pieces employed are good clocks mounted in a fairly uniform temperature the observations 

 should as far as possible be made within the same right ascension limits. If the time pieces 

 are not reliable it is of course better to make the observations simultaneously. Many 

 astronomers will no doubt prefer that the observations should in any case be simultaneous, 

 but the writer's experience seems to point in the opposite direction, under the circumstances 

 named, and as it is a matter which is very closely connected with the permanence of the 

 personal equation, and seeing that the time intervals (the diflerence of longitude in the 

 stations) must in either case be bridged by a clock he prefers to trust one he knows 

 something about and retain other things in their normal condition. 



