252 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



differences of lon«j:itude. In fact, van liemnielen, from the mean of 

 53 abrupt be<rinnin^'s, computed the difference of h)n^itude between 

 Batavia, Java, and (iieenwich, and obtained a vahie differing bj 

 only 9 seconds from the one derived in the usual wa}'. 



On May 8. lUO'i, there occurred a magnetic storm, the abrupt 

 beginning of which coincided with the eruption of Mont Pelee, as 

 nearly as the time of that disastrous outbreak could be determined. 

 This apparent coincidence sujiijested the possibility that the two 

 phenomena might be related. Bauer made a study of this storm in 

 1910, based on the records of 25 observatories, and reached the conclu- 

 sion that it did not begin at the same instant all over the earth, but 

 on the contrary had its origin in about longitude 75° west of Green- 

 wich and traveled eastwardly around the earth, requiring between 

 three and four minutes to complete the circuit. A similar study of 

 another storm of distinctive features supported this conclusion. 



For further evidence on this very important question, Faris made 

 a study of 15 abruptly beginning storms occurring in 1906-1909, 

 using the records of the five magnetic observatories of the Coast 

 and Geodetic Survey, and his results apparently confirmed Bauer's 

 conclusions, the transmission time again coming out about 31/2 min- 

 utes. It was necessary, however, to assume that some of the storms 

 traveled from west to oast, as had been suggested by Bauer. Faris 

 pointed out that i)i-obably the greater part of the error of deter- 

 mining the time of an abrupt beginning is due to the difficulty of 

 selecting the exact point of beginning on the curve. Bauer accord- 

 ingly secured reports from 23 observatories on the same storms 

 which Faris had investigated, making an effort to secure more 

 homogeneous data by means of ti-acings of portions of the curves 

 showing the points selected for measurement. A discussion of these 

 more accurate data failed to confirm the conclusions drawn from 

 the earlier investigations. Kodes also nuide a study of a number 

 of abrupt-beginning storms between ]910 and 1921, and found no 

 evidence of storm propagation of the character suggested by Bauer, 

 but he did find some indications tliat a terrestrial magnetic storm 

 may begin at the "front meridian," that portion of the earth, that 

 is, which "would be the first to meet a stream of electrified particles 

 coming from the sun, occurring later on both sides of that meridian. 



The close correlation between (he occurrence of nuignetic storms 

 and auroras and the presence of large spots on the sun naturally led 

 to attempts to trace a causal relationship. It was soon seen that a 

 direct nuignetic effect by the sun was out of the question. With the 

 development of the idea of curi-ents of electricity being the cause of 

 the earth's nuignetism. different forms of electric discharge emanat- 

 ing from the sun were successively ])ut forth as the cause of the 

 observed terrestrial phenomena, the theories advanced keeping pace 



