164 CARNEGIE INSTITUTION OF WASHINGTON. 



(14) Covellite: A singular case of chromatic reflection. H. E. Merwin. J. Wash. Acad. 



Sci., 5, 341-344 (1915). 



The mineral covellite, CuS, in the finest powder is deep blue. Thin flakes 

 (less than 0.001 mm.) in transmitted light are green. Immersion in liquids 

 of high refraction causes covellite to reflect red or purple from obliquely inci- 

 dent white light. Absorption is not strong enough to account for the color 

 effects. The refractive index, w, for red is less than 1, for yellow 1.5, for blue- 

 green 2. Oblique reflection from such a material in air gives a preponderance 

 of blue; in highly refracting liquids red and orange predominate because of 

 total reflection. The character of the reflection from sections normal to the 

 cleavage shows that e is greater than w for all colors. 



(15) Mechanical strain and thermoelectric power. Walter P. White. Phys. Rev., 6, 



234-236 (1915). 



It is difficult to specify definitely either the amount or character of the 

 changes due to permanent mechanical strain in metals; it is even uncertain 

 what they consist in, and yet their effect upon the thermoelectric power may 

 throw light upon the important and obscure questions which exist regarding 

 the relations of electricity to metals. A few observations upon this effect 

 are therefore given. 



In general, the effect of alloying a metal is to lower the thermoelectric 

 power. The effect of hardening by permanent strain (bending, drawing 

 through dies, etc.) either a pure metal or an alloy is to raise the power. Hence 

 hardening has an opposite effect to alloying, though of less amount. Plati- 

 num alloys, however, are most of them higher than platinum. And the alloy 

 of platinum containing 10 per cent of rhodium (and probably other platinum 

 alloys also) is lowered by mechanical hardening, so that here also alloying and 

 hardening have effects opposite to each other. Pure platinum, however, is 

 changed in the same direction by hardening and (usually) by alloying. One 

 probable result of this opposite behavior of pure platinum and its alloys is 

 that there are alloys of such compositions that hardening has no effect on their 

 thermoelectric power. 



The thermoelectric effects of temporary strain (strain below the elastic 

 limit) have no perceptible relation to those due to permanent strain, or harden- 

 ing. In fact, hydrostatic pressure and tension produce, in a number of cases, 

 opposite effects. Hydrostatic pressure, like hardening, usually raises the 

 thermoelectric power, whence it follows that tension has the opposite effect to 

 hardening. But even this is not true for all metals. Manganin, which is 

 exceptional for both tension and pressure below the elastic limit, is raised, 

 like most other metals, by hardening, while platinum-rhodium is not excep- 

 tional for temporary strains. 



A certain sample of constantan wire, hardened by drawing from 2.5 to 0.25 

 mm. diameter, was softened by careful annealing near 800°. Another portion 

 of the same, annealed near 300°, remained mechanically rather hard, but was 

 as soft, i. e., as low, thermoelectrically as the other. A metal, therefore, may 

 be in a state of relative mechanical hardness luithout possessing the corresponding 

 thermoelectric condition. The same effect has been observed with platinum, 

 which is generally annealed perfectly at a dull red heat, though not then so 

 soft mechanically as after a heating to 1300° or so. In one instance, however, 

 the thermoelectric power was lo\yered by a high heat after it had become 

 apparently constant by heating to a dull red. 



Even in a short communication upon thermoelectricity, it seems necessary 

 at the present time to specify what convention is used regarding the sign of 

 the thermoelectric power. The best authorized notation, used from the dis_ 



