232 



SCIENCE 



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



account of temporary reversals in deflection 

 which are lost sight of in the general deflec- 

 tions lasting for an indefinite period. 



It will be noticed that the pendulums show 

 greater response to pressure conditions during 

 the fall and winter months than during the 

 spring months. This is to be expected, inas- 

 much as barometric maxima and minima are 

 best developed during fall and winter. The 

 records for the summer months were not ex- 

 amined critically on this account. The study 

 thus far has been entirely qualitative; quanti- 

 tative work has been found unsatisfactory 

 owing to the lack of a recording device which 

 shall obviate the running together of the 

 hourly lines at the very frequent times of ex- 

 treme deflection. 



The causes of the movements here described 

 are obscure. Many suggestions regarding the 

 causes of similar movements elsewhere have 

 been made, but no one of them is corroborated 

 as yet by sufiiciently wide-spread observation, 

 to warrant its being fully accepted. It would 

 seem that causes which may be operative over 

 long distances must be assumed, for the 

 pendulums at Cambridge show distinct move- 

 ments in sympathy with barometric maxima 

 and minima when these are still very far dis- 

 tant from the station. 



The possibility of using horizontal pendu- 

 lums in forecasting on windward coasts has 

 been suggested by Mr. F. Napier Denison, of 

 the Meteorological Office at Victoria, B. C. 

 If, as in the case of the Harvard station, hori- 

 zontal pendulums in general announce the ap- 

 proach of various pressure conditions in 

 advance of the barometer, the use of simple 



instruments of this type in situations where 

 maps of weather conditions to windward are 

 not available, might lead, especially in the 

 latitudes of the prevailing westerly winds and 

 cyclonic storms, to valuable results. 



B. M. Vaeney 

 / Habvaed University 



a simple and efficient lecture galvanom- 

 eter arrangement 



In view of the extensive use to which the 

 lecture galvanometer is nowadays put in phys- 

 ical and other laboratories, I have been in- 

 duced to describe a particularly simple ar- 

 rangement which has been thoroughly tested 

 and whose performance leaves little to be de- 

 sired. 



In this arrangement a firm tripod, supported 

 by a convenient shelf on one wall of the lec- 

 ture room, carries a 90° arc lamp clamped by 

 a right-angle piece to its vertical rod. The 

 lamp is mounted with the positive carbon 

 vertical, and its luminous tip, the source of 

 light, uppermost. On a wall bracket a converg- 

 ing lens with its axis vertical is mounted 

 about a meter above the arc. The galvanom- 

 eter, a D'Arsonval instrument with plane 

 mirror, is mounted on a wall sheK with its 

 mirror, A, about 0.4 meter above the lens and 

 about 0.1 meter nearer the wall. A second 

 and larger plane mirror, B, is mounted with 

 universal adjustments at the edge of the gal- 

 vanometer shelf. It is fixed vertically above 

 the lens in a horizontal plane a little below A. 

 A scale with 2-ineh divisions is mounted 

 horizontally near the top of the wall opposite 

 the galvanometer about 9.5 meters away. The 

 galvanometer terminals are permanently con- 

 nected with binding posts on the lecture table. 



When the optical adjustments have been 

 made, light from the tip of the positive car- 

 bon, converged by the lens, falls upon the 

 mirror B and then upon the mirror A, which 

 reflects it to the scale. At the center of 

 the scale a round and brilliant image of the 

 luminous carbon tip is formed. Focal ad- 

 justments can be made by moving the lens 

 vertically on its bracket, or the lamp vertically 

 on its rod; and the position of the image on 



