230 



SCIENCE 



[N. S. Vol. XXXI. No. 789 



cause or location, except in direction, of the 

 gegenschein, but it seems not improbable that 

 it may be more distinctly visible during the 

 passage of the earth through the luminous 

 particles of a comet's tail, and therefore it 

 should be studied, at the proper time, with 

 the greatest care by those in the habit of ob- 

 serving it. 



18. The Auroral Line. — Arrhenius" says : 



Whichever way we turn the spectroscope on a 

 very clear night, especially in the tropics, we 

 observe this peculiar green line. (The so-called 

 auroral line.) It was formerly considered to be 

 characteristic of the zodiacal light, but on a 

 closer examination it has been traced all over the 

 sky, even where the zodiacal light could not be 

 observed. 



Evidently the source of this line is not defi- 

 nitely known, but, conceivably, it may be ren- 

 dered more brilliant by the passage of the 

 earth through the tail of a comet, and there- 

 fore it would be well for some favorably 

 situated observer carefully to measure its 

 brilliancy on several consecutive nights, so 

 selected as symmetrically to overlap the calcu- 

 lated date of our supposed passage through the 

 tail of Halley's comet. 



The most promising, in this connection, of 

 the above phenomena are, in the author's 

 opinion, those designated as a, h, c, d, 5, 6, 

 9, 10, 13, 16 and lY. 



The above is not claimed as a complete list 

 of the phenomena that may be associated with 

 a comet, but it is hoped that they, together 

 with others that they may suggest, will soon 

 give us a better understanding of comets in 

 general and of Halley's in particular. 



W. J. Humphreys 



Mount Weather Obsebvatoby, 

 Bluemont, Va. 



' SPECIAL ARTICLES 



SOME LONG-PERIOD DEVIATIONS OF THE HORIZONTAL 

 PENDULUMS AT THE HARVARD SEISMO- 

 GRAPHIC STATION 



The studies of Omori, Milne, Denison and 

 many others, on the movements of horizontal 

 pendulums due to other than seismic or 



'"'Worlds in the Making." p. 116. 



microseismic causes, suggested a similar 

 study of the movements shown by the pair of 

 Bosch-Omori instruments at the Harvard sta- 

 tion. These pendulums, which stand at right 

 angles to each other on the meridian and par- 

 allel of the station, record through small 

 tracers on sheets of smoked paper carried by 

 drums that complete a revolution once in an 

 hour. The drums travel laterally, causing 

 each hour's record to appear as a single line 

 spaced about an eighth of an inch from its 

 neighbor on either hand. A complete day's 

 record, undisturbed by seismic or other move- 

 ments, appears as a series of twenty-four 

 parallel lines. Any long-period deviations of 

 the pendulums, therefore, are shown by a 

 crowding of these lines toward one side of the 

 sheet or the other. 



The study was made to determine whether 

 or not solar or cyclonic and antieyelonic con- 

 ditions affect the pendulums, as has been 

 suggested. Lack of time prohibited an investi- 

 gation of tidal and other effects, except so far 

 as to prove them entirely subordinate to the 

 main controls. The records were examined 

 for the months of April, May, October, No- 

 vember and December, 1908. The pendulum 

 standing on the meridian of the station (the 

 east-west component, so-called) is most sensi- 

 tive, in the matter of long, non-periodic move- 

 ments, to forces applied due east or west of the 

 station. The same is true of the north-south 

 component in reference to forces applied on 

 the north or south. 



Two types of deflection are shown by each 

 component : 



The E.-W. Component: Type 1. — A diurnal 

 deflection. This is indicated by a more or less 

 strong tendency of the pendulum to move east 

 during the forenoon and west later in the day. 

 It begins about sunrise, the more or less steady 

 easterly travel dying out about noon and later 

 becoming a westerly travel which often lasts 

 well into the night. This type of deflection 

 never persists from one twenty-four hours into 

 the next; it occurs only on days when the sun 

 shines, and is best shown on the least cloudy 

 days. When the diurnal quality of the ther- 

 mograph curve is most marked, the pendulum 



