MOUNT WILSON SOLAR OBSERVATORY. 



259 



as more plates are used the averages fall into a series progressively 

 increasing with length of exposure. 



For counting, a reseau having 44 by 56 squares was used, each square 

 being 30 seconds on a side. The negative was observed directly under 

 a power of 12. The number in each square was noted on squared 

 paper and the vertical and horizontal columns were added and plotted. 

 The N.-S. (horizontal columns) curve in every case is more pointed 

 and has a higher maximum than does the E.-W. curve, though the 

 areas in each case are identical. From the series of short-exposure 

 curves of the same cluster in which stars are counted clear to the center, 

 a general curve is being deduced for each cluster that may be applied 

 to those plates in which the center is burned out and the total number 

 of stars thus obtained. All the photographs are on Seed 23 plates. 



The following results for Messier 13 are typical and indicate the 

 character of the results with increasing exposure time: 



DIRECT DETERMINATION OF PARALLAXES. 



For the parallax work, 282 plates have been taken by Mr. van Maanan 

 with the 80-foot focus combination of the 60-inch reflector. Most of 

 the plates were given two exposures to the same field, and on six of 

 these the distances of the images were measured to obtain an idea of 

 the accuracy attainable in the final results. The probable error of a 

 distance measured in two positions of the plate is ±0':f 030, when all the 

 stars on the plate are included, and 0''023 when only those near the 

 center are used. These figures indicate a very satisfactory degree of 

 precision, but some months must elapse before a wholly reliable opinion 

 can be expressed as to the accuracy. At present provisional parallaxes 

 are available for two stars : 



Boss 6097 T = +0':094 ± OrOlO (4 plates, 2 exposures each). 

 5519 TT = —0.039 =t 0.020 (6 plates, 2 exposures each). 



