proceedings: philosophical society 301 



phic method for the determination of the average interval between depart- 

 ures from the mean greater than a given departure. In determining the 

 absolute winter minimum below which the winter temperature would 

 not fall on the average oftener than once in thirty years, a problem 

 which arose in connection with the investigations of the methods and 

 cost of heating greenhouses, a curve was constructed for which the 

 abscissas are departures from the mean expressed in terms of the stand- 

 ard deviation, and the ordinates represent the reciprocals of the pro- 

 bability of departures greater than a given amount. Thus if the depart- 

 ure in question is an annual event, the ordinates of the curve represent 

 the average interval, in years, between successive occurrences of the 

 event. Periods of observations of meteorological phenomena are too 

 brief to give very reliable results from the method outlined. The 

 method is also applicable only to variables represented by normal fre- 

 quency curves. In the case of 569 stations the actual number of spring 

 frosts after the calculated date be3^ond which frost should occur on the 

 average only one j r ear in ten gave no unexpected frosts in 73 per cent 

 of the stations and only one unexpected frost in 21 per cent of the sta- 

 tions. Thus in 94 per cent of the stations there was either no unex- 

 pected frost or only one. The mean of the standard deviation for these 

 stations was calculated from an average of about 23 observations. 

 The paper was discussed by Messrs. Woodward and Humphreys. 



The 772d meeting was held on Thursday. April 20, 1916, at the Cos- 

 mos Club; President Briggs in the chair, 130 persons present. 



The evening was devoted to an address by Dr. R. A. Millikan on 

 Some recent aspects of the radiation problem. Partly as the result of an 

 experimental situation and partly because of a theory, Planck's h first 

 made its appearance in 1900 in connection with the development of the 

 laws of black-body radiation. Since then it has unexpectedly revealed 

 itself (2) in the domain of specific heats, (3) in that of corpuscular emis- 

 sion under the influence of light and X-rays, (4) in the domain of spec- 

 troscopy, both of light and of X-rays, and (5) in the general radiation 

 which is stimulated by the impact of corpuscles against the atoms of 

 matter. 



This is an extraordinary experimental situation which has not yet 

 been interpreted in the light of any consistent theory. After the pre- 

 sentation of the facts which have recently come to light in connection 

 with the last three of the foregoing fields, it was pointed out that the 

 work of Duane, Hunt, and Hull seems to permit of a real advance in 

 theory in that it appears to show that the h which is found in connec- 

 tion with the general X-ray radiation, and presumably in connection 

 with black-body radiation has nothing to do with the natural periods 

 of the atomic constituents of the radiating bodies as heretofore as- 

 sumed, since it is quite independent of the nature of these bodies. It 

 appears to be rather a property of the ether pulse which is generated 

 by the stopping of an electron. 



