220 



SCIENCE. 



[Vol. VIII., No. 187 



try, and was for a time most actively pursued 

 here, culminating in those beautiful photographs 

 of the moon taken by Rutherfurd, as well as pho- 

 tographs of several double and multiple stars, and 

 of the clusters Praesei)e and the Pleiades, He 

 told how Rutherfurd constructed a micrometer 

 measuring engine, and obtained the first measures 

 of the distances and position-angles of stars upon 

 photographic jplates, and how the work was re- 

 ceived with considerable skepticism abroad. The 

 speaker then described his own continuation of 

 this same kind of work at Cordoba, and stated 

 that he had brought home plates whose measure- 

 ment would take a lifetime. Dr. Gould thought 

 that he had the records of many 11th magnitude 

 stars on his plates, the first photographs of such 

 faint stars. Few of the plates were yet measured, 

 and he was becoming solicitous about obtaining 

 the necessary funds to proceed as rapidly as pos- 

 sible with this measurement, as he had detected a 

 tendency, in some of the plates, of the collodion 

 film to become detached from the plates. 



A paper by Mr. E. F. Sawyer, entitled ' Some 

 account of a new catalogue of the magnitudes of 

 southern stars,' was presented. Mr. Sawyer has 

 been observing the relative magnitudes of all the 

 stars between the equator and — 30°, using an o^^era- 

 glass with the stars slightly out of focus, and era- 

 ploying Argelander's method. Dr. Gould paid a 

 high compliment to Mr. Sawyer's work, as did also 

 Mr, Chandler. 



A paper by Dr. Elkin, of the Yale college ob- 

 servatory, upon ' A comparison of the places of 

 the Pleiades as determined by the Konigsberg and 

 Yale college hehometers,' was presented by Pro- 

 fessor Newton. The results given were provisional ; 

 but they show unquestioned change of position 

 with reference to ?; Tauri since 1840. Most of the 

 brighter stars of the group, as shown by Newcomb 

 in his catalogue of ' standard stars ' go with ?/ Tauri, 

 but among the smaller stars there are unquestioned 

 departures from this community of proper motion. 



In Monday's session a paper by Professor Abbe 

 created some discussion. The point of the paper 

 was, that, as the force of gravity varied from the 

 equator to the poles, thirty inches of mercury in 

 the barometer indicated a less gaseous pressure, 

 and consequently less density of the atmosi)here, 

 at the equator than thirty inches at the poles, and 

 hence a correction for latitude should be intro- 

 duced in allowing for refraction. He showed that, 

 for the difference of latitude of Pulkowa and 

 Washington, it would make 0".l difference in the 

 refraction at 45° of zenith-distance, and might be 

 sufficient pairtly to account for differences in sys- 

 tems of star declinations which depended upon 

 observations at great zenith-distances. 



The most important paper in the section, and 

 the one that attracted the most attention and dis- 

 cussion, was by Mr. Chandler, of Cambridge, upon 

 ' A comparative estimate of methods and results in 

 stellar photometry.' We have not space to do jus- 

 tice to this valuable and rather revolutionary paper, 

 but we will try briefly to give its gist. Prefacing his 

 remarks with the statement that it had long been 

 known that small differences of stellar magnitude 

 could be determined very accurately by Argelan- 

 der's method of steps, by naked-eye estimates, but 

 that it had been generally siapposed that large dif- 

 ferences could not be accurately so determined, and 

 that the general idea had been that, as soon as 

 photometry came generally into use, and so-called 

 measurement took the place of estimation, a much 

 more accurate scale of magnitudes, depending 

 upon a true geometric light-ratio, would at once 

 take the jjlace of the old, the latter becoming ob- 

 solete, Mr. Chandler took for his text the general 

 statement that instrumental photometry had thus 

 far proved a failure ; that is, it had not developed 

 a more uniform scale of magnitudes than Argelan- 

 der's, nor had the accuracy of individual deter- 

 minations been increased, but they were, on the 

 contrary, far more uncertain than the old differ- 

 ential naked-eye estimates. These statements he 

 proceeded to back up with a convincing ari'ay of 

 well-digested results, of which we can only give 

 the briefest summary : I''. For stars of Argelan- 

 der's scale between magnitudes 2 and 6, the pho- 

 tometric catalogues of Seidel, Peirce, Wolf, Picker- 

 ing, and Pritchard differed among themselves as 

 much (or more) in their measures of what Arge- 

 lander called a difference of one magnitude, as 

 they did in their measures of his successive magni- 

 tudes. 2°. Their average values of the logarithm 

 of the light-ratio (we will call it simj)ly light-ratio 

 hereafter, for brevity) for one of Argelander's 

 magnitudes between 2 and 6, ranged between ,30 

 and .38, about .35 for the mean of all the above- 

 mentioned catalogues. 3°. Between magnitudes 6 

 and 9 of Argelander, the catalogues of Rosen and 

 Ceraski averaged about ,35 for the light- ratio, 

 while Pickering's late results with his large 

 meridian-photometer gave (between magnitudes 6 

 and 8.5) .48 instead of ,35 for this ratio. 4". To show 

 the discrepancies in another way, assume a com- 

 mon light-ratio of .35 for all the photometers, and 

 that their scales agree at magnitude 6, Then, 

 for stars of the second magnitude, they will differ 

 by 0.8 of a magnitude. Tfifet is, at a distance of 

 four magnitudes away from where they agree, one 

 photometer will say that the same star is twice as 

 bright as another will, 5°. To test the uniformity 

 of the different scales, all were referred to the 

 average scale of all the photometers, and it was 



