Makcii 1, 1895.] 



SCIENCE. 



235 



score of periods have been determined from 

 25.5 to 27.5 days, depending somewhat 

 upon the data employed and upon the me- 

 thod of its manipidation. It is very certain 

 that, if any one will take the ' horizontal 

 force,' for example, and arrange the observa- 

 tions in intervals of 2() days (the best thus 

 far found) he will quickly find that, judg- 

 ing from the disturljed days, there is abso- 

 lutely no fixed interval. These disturbed 

 days would seem tlie verj- best material for 

 such studies, as they are very definite. 

 These days will occur for three or four ro- 

 tations most beautifully, but after that the 

 distui-bance disappears and no more will 

 appear along that line for a score of rota- 

 ti<ms. 



In the .same May one will very quickly 

 find, in using the data and leaving out the 

 disturbed days, that there is absolutely no 

 recurring period of 2(5 days or any fraction 

 of that interval. Sometimes by grouping 

 ten rotations one will find a fairly good 

 fluctuation, but the very next group of ten 

 rotations will make ' hedge podge ' of the 

 pre^aous group. This woidd seem an ex- 

 tremelj' important point to settle, as months 

 Jiave been devoted to fruitless efibrts in 

 trying to determine such a period. 



3. The fluctuations under (c) above are 

 simultaneous over the whole earth, as has 

 been shown by the records at Batavia, 

 India, Los Angeles. Cal.; St. Petersburg 

 and Tiflis, in Russia ; Vienna, Austria ; 

 ■Washington, D. C, and Zikawei, China. 

 One is struck at once by the wonderful 

 regularity of these fluctuations over the 

 whole Northern Hemisphere. Making al- 

 lowance for the difterence of time and for 

 disturbed days, the fluctuations an; found to 

 he exactly the same at each station, and the 

 record at a single station will answer per- 

 fectly for (comparison with anj' supposed 

 related meteorologic jthenomenon. 



4. After thirteen years of study and care- 

 ful discussion I am satisfied that the pressure 



of the air, or perhaps the fluctuations of the 

 dew jioint, arc by far the best to use for de- 

 termining a possible connection with mag- 

 netism. I am also perfectly satisfied that, 

 except in the cases specified above, there is 

 no direct relation between magnetic and 

 meteorologic phenomena, and this is also 

 the outcome of the exhaustive studies in 

 England and on the continent. I am also 

 satisfied that thei-e is an indirect relation, 

 but the phenomena are so extremelj^ com- 

 plex that it has proven impossible to deter- 

 mine it up to the present. 



5. In all studies of this character, and in 

 all attempts at determining coincidences 

 between such phenomena, one will always 

 find a most valual)le check by cutting up 

 the long list ol' rotations into groups of 7, 

 10 or l-t rotations in each. If these sepa- 

 rate gi'oups do not show a thread running 

 through them, or fluctuations common to all 

 and continuall}- recurring, he may be satis- 

 fied that there is nothing in it. There is a 

 peculiar and well-nigh unaccountable fasci- 

 nation in arranging and summing grouj)s of 

 figures in the hope that something may 

 come, but continuous effort will show that 

 there is something back of it all which is 

 not understood, and no headway can be 

 made bj' direct comparisons. 



H. A. Hazen. 



Febriary 1, 1895. 



SIMILAR INVENTIONS IN AREAS WIDE 

 APART. 



As a contribution to the much (lisi)uted 

 question of the occurrence of similar inven- 

 tions in areas wide apart, I desire to call the 

 attention of readers to the device for weav- 

 ing of which I have found abundant ex- 

 amples in the Pueblo country, in the New 

 England States, and in Finland. 



The apparatus consists essentially of a 

 small rectangular frame-work in which are 

 placed a series of jierpendicular slats i)erfo- 

 rated in the middle. It has the appearance 



