458 



SCIENCE. 



[N. S. Vol. XXIII. No. 586. 



in December, 1904, a brief aceoimt was 

 given by the writer of methods and results 

 in a study of faint variable stars recently 

 undertaken at Plarvard College Observa- 

 tory. The most remarkable regions as yet 

 examined are the Nebula of Orion and the 

 two Magellanic clouds. These were men- 

 tioned a year ago, when it was reported 

 that the numbers of variables in the large 

 and the small Magellanic clouds were one 

 hundred and fifty-two and fifty-seven, re- 

 spectively. In the autumn of 1904, a series 

 of sixteen plates was taken at Arequipa, 

 covering the region of the small Magellanic 

 cloud. These ai'rived at Cambridge in 

 February and were examined with great 

 interest, as it was expected that the number 

 of variables would be considerably in- 

 creased. The result of the examination, 

 however, exceeded all anticipations. So 

 many were found that it was April before 

 the preliminary study of the plates was 

 completed and it could be announced that 

 nine hundred new variables had been dis- 

 covered. The number has since been in- 

 creased to nine hundred and seventy, and 

 new ones are still being discovered from 

 time to time. Measurements of the posi- 

 tions of all these, selection and measure- 

 ment of a large number of sequences of 

 comparison stars, and the determination 

 of a provisional light-scale, have necessarily 

 consumed much time, so that it is only 

 within a few weeks that it has become pos- 

 sible to begin actual observation of the 

 brightness of the variables. Light-curves 

 thus far deduced are of well-defined types. 

 Periods of many of the variables appear to 

 range from one and a quarter to four days. 

 A number, mainly among the brighter ones, 

 have periods of from one to two weeks, 

 while very few appear to have long periods. 

 A preliminary examination of recent photo- 

 graphs of the large Magellanic cloud shows 

 that it resembles the region now under dis- 



cussion and it is probable that a very large 

 number of variables will be found to exist 

 within its limits. Preparations for meas- 

 uring the positions and brightness of these 

 are already being made. 



The question as to whether there are 

 other regions containing such remarkable 

 groups of variable stai-s, is of the greatest 

 interest. Not less important is the general 

 problem of the distribution of these fainter 

 variables. An exploration of new regions, 

 therefore, is being undertaken as time per- 

 mits, and even negative results, though less 

 immediately interesting, are valuable as a 

 contribution toward an ultimate solution 

 of the problem. 



Amherst Eclipse Expedition to Tripoli, 



1905: David Todd. 



Instruments installed on the rigid roof- 

 terrace of the British Consulate-General. 

 Six departments of observation: 



1. Observations of the geometric con- 

 tacts. 



2. Coronal photography with a twelve- 

 inch Claeey lens of nine feet focus, photo- 

 graphically corrected. Eighteen exposures 

 with semi-automatic movements. Baily's 

 beads also photographed before second con- 

 tact. My modified form of Burckhalter 

 revolving occulter was employed on the 

 corona. Corona photographed to 30'. 



3. Duplex Clark lenses (photographic) 

 of three inches aperture and eleven feet 

 focal length, for long exposures on the 

 circum-solar stars and the outer coronal 

 streamers. No intra-mercurial planet re- 

 vealed on 14 x 17 plates of the highest 

 sensitiveness, stars to seventh and eighth 

 magnitude. 



4. A three -and-one-half -inch Goerz 

 doublet of thirty-three and one half inches 

 focus, attached to one of the automatic 

 movements used on my previous expedi- 

 tions of 1896, 1900 and 1901, secured 63 

 fine pictures of the corona during the 186 



