proceedings: philosophical society 113 



observ'ed during the total eclipse, and the experiences encountered en- 

 route to Liberia and in Liberia itself. 



Totality lasted at Cape Palmas about 6 minutes and 33 seconds, 

 longer than at any other station in the belt of totality. The general 

 indications, as the eclipse occurred during the rainy season, were that 

 Cape Palmas would not be a suitable station for the astronomer. How- 

 ever, for the purpose of the Department's investigations, it did not 

 matter whether there was a clear sky or not, for a magnetic effect will 

 pass through any layer of clouds. It happened, however, that in spite 

 of general expectations, the weather was clear, and this now for the 

 third time, whereas certain parties at other stations, which appeared 

 more favorable according to past meteorological records, were un- 

 fortunate. The observation program included magnetic and electric 

 observations, meteorological observations, shadow-band observations, 

 times of contacts, and photographs such as could be obtained with 

 small kodak cameras. This comprehensive program was carried out 

 successfully, excepting the atmospheric-electric work which, on account 

 of the deterioration of the dry-cell batteries purchased in England, 

 had to be abandoned. Although three observers had been stationed, 

 no shadow bands were observ^ed this time, even greater precautions 

 having been taken than at Corona during the eclipse of June 8, 191 8, 

 where they were observed. 



The eclipse of May 29, as observed at Cape Palmas, was not nearly 

 as dark, in spite of its long duration, as the much shorter one of June 

 8, 1 91 8, at Corona, Colorado. There was a marked difference in light, 

 both as seen visually and as shown by the. photographs, between the 

 inner corona and the outer extensions. The large red prominence 

 was a startling object. 



Clear indications were had with regard to a magnetic effect in ac- 

 cordance with the results obtained at previous solar eclipses. 



There was a steady slight decrease in temperature from 12*^ G.M.T., 

 0.7 minute after the first contact, to 12.7^ G.M.T., and then a more 

 rapid decrease until 14^ G.M.T., when the minimum temperature of 

 79.4° F. was reached. This time (14.^) was approximately 0.4'' later 

 than the middle time of totality. The increase in temperature after 

 14'^ was rapid, the maximum of 82.7° F. being reached at 14. 9*^ G.M.T. 

 The hydrogram for May 29 showed the following effect: the humidity, 

 which was 71 per cent at 12*^ G.M.T. , steadily increased to 78 per cent 

 at 14** G.M.T. There was a more rapid decrease from 14'^ G.M.T. 

 to 15** G.M.T. when the humidity was 66 per cent. The maximum 

 humidity, therefore, occurred at 14^ G.M.T. or approximately 0.4 

 hour later than the middle time of totality. The barogram showed 

 nothing marked during the time of the eclipse. 



D. M. Wise: The total solar eclipse at Sobral, Brazil, May 2g, 1919. 



At Sobral, Brazil, the Department of Terrestrial Magnetism carried 

 out a program of magnetic and atmospheric-electric observations very 



