OF ARTS AND SCIENCES. 221 



by increasing the speed of the machine. Following the method of 

 Siemens, the greased disk of the photometer was moved farther and 

 farther from the incandescent platinum, so as to keep the disk uni- 

 formly illuminated, until finally melting occurred. The last reading 

 of the position of the disk gave the illuminating power of the wire at 

 melting. An unexpected constancy in the results was found, so that it 

 seemed worth while to make a longer series of such observations, which 

 was done in the following manner. At one end of the bar of a greased- 

 disk photometer 200 centimeters (78.74 inches) long was placed a 

 circular aperture, through which shone the light of an Argand gas- 

 burner. A platinum-wire lamp was constructed, in which a given 

 length of wire was stretched between two fixed knife-edges, and held 

 by clips, thus allowing of speedy adjustment of the wire. The disk 

 used was furnished with two mirrors inclined at 45° in the usual man- 

 ner, and it was always placed so that the two sides as seen by reflec- 

 tion appeared alike, — a method of measuring much more satisfactory 

 than the one commonly employed with the Bunsen photometer. The 

 length of wire used was measured between the edges when cold, and 

 allowed to expand freely. An attempt was made to use a slit of given 

 length in front of a longer wire, which was abandoned owing to errors 

 arising from inequality in the length of that portion of the wire whose 

 light passed through the slit when the wire expanded. The use of a 

 small weight stretching the wire tight, as suggested by Professor Pick- 

 ering and employed by him in a small lamp, presents some advantages, 

 but was discarded from fear of elongating- the wire and so causing it 

 to break before complete fusion was reached. Melting was brought 

 about by the use of a slide-resistance of German-silver wire in circuit 

 with the platinum wire and a dynamo machine. The variable resist- 

 ance of the slide was lessened until the platinum wire melted, the 

 photometer-disk being meanwhile moved repeatedly, its final reading 

 giving the light emitted by the platinum in terms of the light emitted 

 by the aperture opposite. 



No attempt has been made in these experiments to determine the 

 absolute illuminating power of the wire, but only to ascertain whether 

 ordinary commercial platinum wire has any practical value for use as 

 a primary or a secondary standard. 



Two different sizes of wire were used, whose diameters were .004 

 in. and .006 in. respectively. Wire from two different makers was 

 also used, ordinary wire drawn without special care being taken be- 

 cause of its availability. The length used was always the same when 

 cold, viz. two inches. A very large number of measurements were 



