NATURAL PHILOSOPIIT. 149 



2. That that of the elements of the second kind (the Grove type) 

 auo;ments with the temperature. 



3. That that of the bmee type is independent of the variations 

 of temperature. In order to verify these results, it suffices to 

 oppose pole to pole two elements quite identical, and to place in 

 the circuit a galvanometer, the needle of which points to zero. 

 By heating one of the element-s with suitable precautions, the 

 needle will deviate in a permanent manner, in a sense which 

 varies with the nature of the element. — Comptes Bendus, leb, 22, 

 18G9. 



DURATION OF ELECTRIC DISCHARGES. 



In the last number of " Silliman's Journal " we notice an admira- 

 ble paper by Prof. O. N. Rood on the aljove subject. His appa- 

 ratus, in general arrangement, resembled that used by Wheatstone 

 in his remarkable experiments for determining the velocity of 

 electric conduction, which were, no doubt, the basis of all similar 

 processes for the measurement of extremely small intervals of 

 time which have been since devised. The improvement intro- 

 duced by Prof. Rood consists chiefly in throwing the image of the 

 spark elongated by the rotating mirror, by means of a lens, on a 

 plate of ground or polished glass, and viewing it with a magnify- 

 ing lens. There were, besides, various ingenious devices in the 

 details of structure and arrangement, such as have distinguished 

 other investigations of tliis eminent physicist. As a result of 

 various experiments. Prof. Rood finds, in the first place, that the 

 discharge from a Lej^den jar charged by an induction coil is com- 

 posed of three parts or successive acts, — a white flash, a brown- 

 ish-yellow gleam, and a fainter and more hxsting glow, which, 

 with brass electrodes, is green, and with platina gray in color. 

 The duration of the entire phenomena is from .000017 to .000050 

 seconds, that is, 17 to 50 millionths of a second; the duration of 

 the white and yellow light together, 6 to 8 millionths ; and the 

 duration of the white flash less than 21: hundred millionths of a 

 second. 



ARTIFICIAL COLORATION OF THE ELECTRIC SPARK. 



Mr. E. Becquerel has shown that the electric spark may be di- 

 versely and beautifully colored by being made to pass through 

 saline solutions. If an electrical spark from an inductive appara- 

 tus be made to pass into the extremity of a platinum wire sus- 

 pended over the surface of the solution of a salt, this spark will 

 acquire special coloration according to the chemical composition 

 of the solution traversed. The saline solutions are best concen- 

 trated, and the platinum wire positive. The experiment is read- 

 ily performed in a glass tube. 



Salts of strontia will color the spark red ; chloride of sodium, 

 yellow; chloride of copper, bluish green, etc. 



The light from these sparks, analyzed by the spectroscope, fur- 



13* 



