174 REPORTS OF INVESTIGATIONS AND PROJECTS. 



Professor Kapteyn's Investigations. 



Professor Kapteyn has presented the following account of his work on 

 Mount Wilson. The greater part of his time has been devoted to the further 

 investigation of the selective loss of light in its progress through space. As 

 the investigation published last year (Contributions from the Mount Wilson 

 Solar Observatory, No. 31) pointed to the conclusion that, other things being 

 equal, the redness of stars must increase with their distance, the degree of 

 redness was investigated of all the stars (over 1,400) for which reliable val- 

 ues of all the required data (spectrum, proper motion, photographic and vis- 

 ual magnitude) could be found in the existing literature. 



The result finally arrived at is that the selective loss of light is certainly ex- 

 ceedingly small. The degree of redness, as expressed by the difference, 

 photographic brightness — visual brightness, increases by only i.o per cent 

 (0.009 magnitude) for a distance of 100 light-years. Owing to the smallness 

 of this quantity, it is not surprising that the probable error of the result 

 amounts to 20 per cent of the whole. 



It must be acknowledged that a perfectly satisfactory treatment of the 

 problem requires the solution of another question : Assuming the spectra of 

 two stars of very different total light-power to show identically the same 

 lines, does it follow that the intensities of the continuous spectra are propor- 

 tional throughout their entire length? It should be possible to settle this 

 question satisfactorily as soon as experiments can be made with this particu- 

 lar purpose in view. In the absence of such experiments, however, it can not 

 be denied that some doubt must remain as to the reality and the amount of 

 the present result. 



In consideration of the fundamental importance of the problem, another 

 method of establishing the reality of the selective loss of light is now being 

 tried. A great difficulty of the method employed up to the present time lies 

 in the fact that, owing to the moderate distances of the stars thus far con- 

 sidered, the effect of the absorption can not exceed a few hundredths of a 

 magnitude — an effect so small that it can be clearly detected only in the aver- 

 age of hundreds of stars. On the supposition that space is somewhat uni- 

 formly filled with absorbing or scattering matter, this effect must, however, 

 increase proportionally to the distance. Thus we need only take into con- 

 sideration objects of sufficient distance in order to get effects which will be 

 quite considerable. 



The difficulty is that such great distances can not be measured by any known 

 method. But there are strong reasons for supposing, however, that most, if 

 not all, of the smaller clusters must be at distances of the required order. An 

 attempt has therefore been made to use the enormous light-gathering power 

 of the 60-inch reflector to obtain all the required data for a series of such 

 objects. 



