MOUNT WILSON OBSERVATORY. 255 



sented graphically by a figure showing a section of the stellar system 

 through the center at right angles to the plane of the Milky Way, the 

 results give lines of equal density which are elliptical in form. For a 

 density 1/200 of that near the sun the longer axis lying in the Galaxy 

 is about 8 times the shorter axis towards the pole. 



In another investigation, now nearly ready for press, the accuracy 

 attainable in results such as these has been investigated. The deter- 

 mination of the arrangement of stars in space, as given in Contribution 

 No. 188, rests mainly on a knowledge of (a) the luminosity curve and 

 (6) the curve giving the total number of stars, N„i, for the different 

 apparent magnitudes m. Observations do not give these curves com- 

 pletely but only to certain limits. The question is whether these 

 limits are sufficiently extended for the purpose. 



For the luminosity curve (a) it appears that the limit now reached 

 will be sufficient for the investigation of the arrangement of stars in 

 space so long as our observations do not extend below the thirtieth 

 apparent magnitude. It will therefore satisfy our wants for many 

 years, perhaps many centuries. 



For curve (6), on the other hand, our observational data are not 

 nearly what could be wished. For values of N^ down to magnitude 

 17 visual, which is well within the limit of the Mount Wilson survey 

 of the Selected Areas, the agreement with an error curve is here also 

 remarkably close. If we assume that the error curve represents equally 

 well the total numbers of stars for all magnitudes fainter than 17.0, 

 then the present data will be sufficient for a fairly good determination 

 of the distribution of the stars in high galactic latitudes up to distances 

 well beyond the median (that is, the distance within which half of all 

 the stars are to be found). For a reasonably satisfactory determin- 

 ation of the median in the direction of the Galaxy, however, our 

 knowledge of the values N^ must be extended to stars of about the 

 twentieth magnitude. If we can not assume that the analytical curve 

 derived from the observed values of N^ applies also for m fainter than 

 17, the problem becomes much more difficult, and, in a certain sense, 

 insoluble 



Nevertheless, a very probable limit of distance can be assigned up 

 to which the densities can be calculated, although this hmit is not yet 

 fully determined. In all probability it will be found to be considerably 

 in excess of 10,000 parsecs. Corresponding limits will also be derived 

 for the cases in which the total numbers of stars down to w = 18, 19, 

 20, and 21 are known. 



A first attempt has also been made to determine the distribution of 

 the stars in the Milky Way cloud between /3 and y Cygni. For a 

 reliable determination, however, counts of the stars down to the twen- 

 tieth visual magnitude will be required. 



