Browning and Uhler — Gallium-Indium Alloy. 353 



did not appear to diminish in mass on continued, intense local 

 heating. This globule was subsequently burned in the carbon 

 arc and the spectrogram showed that it was nearly pure gallium. 

 As might be expected from the tabulated boiling points (at 

 76 cm pressure) most of the copper (b. p. 2310° C.) had not been 

 driven off. Not a trace of lead (b. p. 1525° C.) could be 

 detected. The amount of indium remaining was very small. 

 It seems safe to conclude that the boiling point in vacuo of 

 pure gallium is greater than 1600° C. and that prolonged heat- 

 ing in a suitable furnace at a temperature of about 1800° C. 

 would effect a complete separation of indium from the alloy. 



The density of the specimen was found to be 5'91 grams per 

 cubic centimeter at 20° C, the value given in the standard 

 tables being 5*95. The smaller value cannot be accounted for 

 on the ground of contamination with silicon from the walls of 

 the tube because this hypothesis is not consistent with the 

 spectroscopic evidence. The melting point of the globule was 

 not absolutely sharp but the metalconsistently fell through a 

 platinum ring, of smaller diameter than the solid disc, at a 

 temperature of 29*5° C. As recorded by other observers, the 

 metal remained liquid at 0° C. and showed no signs of con- 

 gealing at the expiration of several hours. It may be of 

 interest to note that when, on different occasions, small globules 

 of the fairly pure metal and very impure alloy were partially 

 volatilized in the electric arc, there usually remained on the 

 positive carbon a grayish white, vitreous mass having a defi- 

 nite crystalline structure. 



Notwithstanding the fact that a thorough investigation of 

 the spectroscopic properties of gallium is now in progress, it 

 may not be premature to record the results already obtained in 

 the process of general orientation. The alloy was volatilized 

 in a direct current arc (110 volt mains) with regraphitized 

 Acheson graphite electrodes. The spectrograph contains an 

 excellent concave grating (by Anderson) of 1 meter radius of 

 curvature and is an improved form of the instrument designed 

 and made by one of us in the year 1904.* The o dispersion is 

 such that l mm includes, on the average, 25*25 A in the first 

 order. „ Since the camera covers the interval from 2000 A to 

 7600 A, the identification of ultra-violet lines is facilitated by 

 the superposition on the first order of the second and third 

 orders. The films used were made to order through the kind- 

 ness of Dr. C. E. K. Mees, and they are easily sensitive from 

 2000 A to 7500 A. The wave-lengths given below are highly 

 tentative. The first negative of the alloy (taken January 22, 

 1916) brought out the following facts which have been sub- 



*See "Atlas of Absorption Spectra" by H. S. Uhler and R. W. Wood, 

 Carnegie Institution of Washington, Publication No. 71, May, 1907. 



