134 



EDWARD C. PICKERING. 



hand, variables of long period and large range can be observed to great 

 advantage by Argelander's method. It is only necessary to select stars 

 very slightly brighter and fainter than the variable, and estimate the 

 intervais in gracies, taking great care to identify correctly both the 

 variable and the coraparison stars. When the range is large, the mere 

 staternent, that the variable is slightly brighter than a given star and 

 fainter than another, is nearly as valuable as an estimate in grades. The 

 number of known variables of long period is now so great, that many 

 observers could occupy themselves in making such observations. Each 

 star should be observed by some one at least once a month, and prefe- 

 rably once a week, and observations when the star is visible only in the 

 morning twilight are of spécial value. The greater portion of stars of 

 this class appear to be irregular in their variations, so that observations 

 extending over many years will be required to détermine the nature of 

 their changes. 



When the stars are brighter than the tenth magnitude, the charts 

 of the Bonn Durchmusterung are indispensable in identifying the régi- 

 ons and the comparison-stars. The récent republication of thèse charts, 

 which had become out of print, now brings them within the reach of 

 ail. The size of the charts and the minuteness of the images of the 

 stars, however, render it difficult to use them at the télescope. 

 Accordingly, photographie enlargements on a scale of 1' = 0. 1 cm., 

 hâve beenmadeof the région 3° square surrounding 69 of thèse variables. 

 Copies of thèse enlargements will be given free of cost to experienced 

 observers, who are ready to cooperate in thèse observations, (Circular 

 N°. 53). For stars from the tenth to the thirteenth magnitude, Hagen's 

 Atlas Stellarum Yariabilium, Séries I, II, and III, is equally indispen- 

 sable. 



Such observations as have been described above, together with a 

 vast number of similar observations which have been made during the 

 last half century, would have but little value unless reduced to a uniform 

 System, so that they may be compared with one another. A large part 

 of the photometric work of the Harvard Collège Observatory during 

 the last twenty five years may be used for this purpose. Eather more 

 than a million settings have been made with the meridian photometer 

 at the stations in Cambridge and Arequipa, Peru. The light of ail the 

 stars from the North to the South Pôle, and of the seventh magnitude 

 and brighter, has been determined upon a uniform system of stellar 



