220 ANNUAL REPORT SMITHSONIAN INSTITUTION, 192 8 



was 3.6. We could detect no difference at all in the value of the read- 

 ings lohen the Milky Way was overhead and when it was out of sight. 

 Our error in the mean values of these readings could scarcely be 

 more than 0.1 ion. Even if we double this estimate so as to have a 

 wide factor of safety, we may conclude at least that the Milky Way 

 exerts no influence upon the cosmic rays which it is 3^et within the 

 power of the instruments used to detect, and that this should mean 

 that the rays coming from the direction of the INIilky Way are not 

 6 per cent greater or less than are those coming from the portion 

 of the heavens at right angles to the Milky Way. This is in agree- 

 ment with our preceding less discriminating measui^ements, and also 

 with recent very careful work at sea level by Hoffmann and Steinke, " 

 who can find there no directional effect in cosmic rays at all; but 

 it is at variance with results reported by Biittner ^* and by Kol- 

 horster." 



This present work was, however, done under quite as favorable 

 conditions as have ever been used. It is very important to obtain 

 unambiguous evidence upon this point. No entirely trustworthy con- 

 clusions about the origin of the rays can be drawn until it is settled. 

 As yet, the case for a favored region from which the rays come does 

 not seem to have been established, but more sensitive tests can be 

 made and will be made in the near future. 



OBSERVATIONS IN CALIFORNIAN MOUNTAIN LAKES 



« 



The object of the new group of experiments at Arrowhead and 

 Gem Lakes, begun early in 1927, was to use an increased electroscope 

 sensibility and an increased accuracy in the determination of the 

 electroscope constants, for the sake of introducing greater preci- 

 sion into cosmic ray determinations and placing the whole subject 

 upon a more strictly quantitative basis. 



As already indicated, different observers are still wide apart on 

 the absolute value of the ionization, though a considerable group of 

 us now find it to be between one and two ions at sea level. This, 

 however, can scarcely be called quantitative agreement. But this 

 could scarcely be expected, since no observers except ourselves have 

 thus far been able to determine the zeros of their instruments; so 

 that most reported values of ionizations must be regarded as esti- 

 mates rather than measurements. Our own values suffer from rather 

 large uncertainties in the determination of the capacities of our 

 electroscopes. 



As to mean absorption coefficients, Kolhorster and ourselves are 

 now in reasonable agreement, but no one except ourselves had untijj 



"Steinke, Zeit. f. Phys., 42, 570; 1927. 

 iSButtner, Zoit. f. Geophys., 2, 190; 1926. 

 "Kolhorster, Natiirwissenschaften, 14, 936; 1926. 



