Temperature on the Colour of Pigments. 403 



provided with a micrometer eyepiece reading to 0*0025 millim., 

 were now focused upon these lines, so that when the strip 

 expanded longitudinally, under the heating action of the cur- 

 rent, the increment could be readily and accurately measured. 

 The indications of this platinum thermometer enabled us to 

 calculate the temperatures to which, from time to time, the 

 foil and the film of pigment placed upon it were heated. In 

 these computations we used Matthiesseir's well-kr own formula, 

 in which the length of a platinum wire is expressed as a 

 function of its temperature. This formula, 



l=l (l + -00000851 1 + -0000000035 1 2 ), 



was originally determined between 0° and 100° C, and is 

 strictly applicable only within that range. Beyond its proper 

 interval it doubtless yields false values, and for very high 

 temperatures the correction may be a considerable one. 

 Although the correction-factor is at present unknown, the 

 equation gives a convenient arbitrary scale of temperatures, 

 wdiich possesses the advantage of being readily reduced to the 

 Centigrade system when the proper formula shall have been 

 determined. 



In the preparation of the film of pigment one of the two 

 following methods was employed. In the case of the oxides 

 of zinc and of magnesium the foil was smoked over the flame 

 of the burning metal ; in other cases the ribbon was well 

 moistened with alcohol and the pigment was sifted upon it 

 with a fine cloth, a quantity just sufficient to entirely conceal 

 the metal being thus applied. Upon the evaporation of the 

 alcohol there remained on the platinum a thin adherent layer, 

 which possessed the natural surface of the dry powdered pig- 

 ment, nearly free from gloss or surface-reflexion. 



The instrument by means of which the photometric mea- 

 surements were carried out was a form of polarizing spectro- 

 photometer. The arrangement of its several essential parts, 

 together with that of the apparatus used in connexion with it, 

 is shown in the accompanying diagram (fig. 1). F F is the 

 platinum strip, the mounting of which has already been 

 described. To it were attached wires from a storage-battery 

 of twenty cells, the amount of current, and consequently the 

 temperature of the foil, being controlled by the variation of 

 resistance in the circuit. The letters m m denote the positions 

 of the diamond-lines ; M M are the microscopes focused upon 

 them for the measurement of temperature. The other side of 

 the platinum foil, upon vdrich the coating of colouring-matter 

 is spread, is illuminated by an incandescent lamp, L/, which 

 is supplied from another storage-batterv. Light from this 



2E2 



