86 



MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



Solar intensities during eclipse — Continued. 



In the preceding table no intensities are given for the Violle bulbs between 11.10 a. m 



and 12.10 p. ni., because the fraction -^ — becomes indeterminate in this interval. The 

 ' u' — au 



reduction of the observations with the black and bright bulbs freely exposed is not given, because 



of their approximate character. The former are, however, used in drawing the curve previously 



given under the head " black bulb thermometer." 



The relative intensities are expressed in the same unit u.sed in reducing all the radiation 

 observations, viz, the intensity in each series at 12.00 May 1 adopted as 1. In order to express 

 graphically the observations, the following chart is given. In it the relative intensities obtained 

 with the conjugate thermometers are represented by © and the curve drawn through them. 

 A query, ?, at the side of the observation indicates that the sun was in a cloud at the time, which 

 has been taken account of in drawing the curve. The dotted line is used to connect the observa- 

 tions with the Violle bulbs, but no curve is drawn, since it is evident that its position would 

 be quite conjectural, on account of the slow recovery of the instruments from the effects of the 

 passing clouds. The broken curve at the top of the chart shows the usual form of the curve 

 of intensity given by the conjugate thermometers, reproduced here for the sake of comparison. 



An examination of the observations and the results of the computation leads to the following 

 conclusions : 



1. The readings of all the instruments- were essentially the same at the close of totality, and 

 agreed with the observed air temperature. This indicates that during the total phase no heat 

 was received by the earth from the atmosphere, as far as these instruments allowed its measure- 

 ment. It need hardly be said that the radiation instruments are not intended to show the efiect 



