436 Professor Elias Loomis. 



in these middle latitudes. Also from the daily observations 

 made by Messrs. Herrick and Bradley at New Haven during 

 seventeen years, he concluded that auroral displays in the 

 middle latitudes of America are generally accompanied by an 

 unusual disturbance of the sun's surface on the very day of 

 the aurora. The magnetism of the earth, the Aurora Borealis, 

 and the spots on the sun, have thus all three a causal connec- 

 tion, and apparently that connection is closely related to the 

 conjunctions and oppositions of certain planets. 



Shortly after the publication of this memoir, Professor 

 Lovering published his extensive catalogue of auroras. A 

 further discussion of the periodicity of the auroras was under- 

 taken by Professor Loomis and published in 1873. In this 

 he made use of all the auroras recorded in Professor Lovering' s 

 catalogue. They confirmed his previous conclusions, only 

 slight modifications being required by the new facts presented, 

 and by their more systematic collation. 



In these papers, as in most of his papers upon other sub- 

 jects, Professor Loomis was ever intent upon answering the 

 questions : What are the laws of nature % What do the phe- 

 nomena teach us ? To establish laws which had been already 

 formulated by others, but which still needed confirmation, was 

 to him equally important with the formulation and proof of 

 laws entirely new. 



Let us now turn to another important line of Professor 

 Loomis's work, — Astronomy. As I have said, he was early 

 interested in the shooting stars. In October, 1834, he read a 

 paper before the Connecticut Academy of Arts and Sciences 

 upon this subject, probably in substance that which was shortly 

 afterward published in this Journal. The published paper is 

 principally a restatement of the observations made in Germany 

 in 1823 by Brandes in concert with his pupils for determining 

 the paths of the stars through the atmosphere, together with 

 methods of computation. From the results of Brandes' 

 observations, however, he deduces an argument for the cosmic 

 character of the shooting stars. One month after reading this 

 paper to the Connecticut Academy he engaged in similar con- 

 certed observations with Professor Twining, who was then 



