PHYSICS. 477 



centimeters long, passing just through the cork and narrowed above. 

 The subterranean canals bringing water to the bottom of the geyser are 

 represented by the other tube, narrower than the first, which descends 

 into the globe where it is a little recurved upward. The upper end of 

 this tube passes from the bottom of a lateral jar half filled with water, 

 which is about on a level with the top of the vertical tube. The whole 

 being filled with water to a level a little below the narrowed point of the 

 large tube, it is only necessary to heat the globe with a Bunsen burner 

 to obtain projections of boiling water to a height of 2 or 3 meters. 

 {Wied. Ann., V, xv, p. 173; J. Phys., June, II, i, p. 290.) 



Yiolle has measured with great care the boiling point of zinc. De- 

 ville and Troost some years ago gave it at 1048°, while Becquerel's de- 

 termination makes it 932°. In the course of other experiments, Violle 

 observed that silver, which melts at 954°, did not fuse in zinc vapor, 

 lie then undertook to determine the boiling point of zinc by means of 

 a special apparatus, and has found it to be 929°.G, or practically 930°. 

 {Comptes Bendus, March, xciv, p. 720.) In a note, Troost says that this 

 value was redetermined by Deville and himself, and fixed at 942° as a 

 mean of 27 experiments. {Comptes Bendus, March, xciv, p. 788.) 



Troost has determined the boiling point of selenium with a view to 

 use this substance for determining vapor densities by the method pro- 

 posed by Deville and himself in 1880. The temperature was measured 

 by air thermometers, with both porcelain and glass bulbs, and was 

 found to be 664° to 666°, or 665° as a mean. The author proposes se- 

 lenium vapor, therefore, because glass bulbs can be used in it without 

 softening. [Comptes Bendus, June, xciv, p. 1508.) 



3. Conduction and Badiation. 



Thoulet has determined the thermic conductibility of rocks, defining 

 the coefficient of conduction to be the quantity of heat which enters or 

 leaves a homogeneous wall of indefinite extent during a unit of time, 

 traversing a unit of surface, when the thickness is unity, and the differ- 

 ence of temperature on the faces is 1° C. He has given the preliminary 

 results obtained with his apparatus, which show the coefficient for glass 

 to be 2.66, for wrought iron 10.26, and for Vizille anhydrite 4.56. (Ann. 

 Chim. Phys., June, V, xxvi, p. 261.) 



Gratz has investigated the conductivity for heat of gases and its re- 

 lations to temperature. He concludes : (1) that heat conduction in the 

 gases — air, hydrogen, and (at low temperatures) carbon dioxide, — consists 

 in the transference of progressive energy only, intramolecular energy 

 contributing immeasurably little ; (2) that the relation of heat conduc- 

 tion to temperature is found by experiment to be such as Clausius's 

 theory requires ; (3) that all results for gases and vapors showing di- 

 vergences from the values calculated from theory are without eviden- 

 tial force since they give only the apparent heat-conducting .power in 

 consequence of the absorption of radiant heat ; (4) that the divergence 



