248 CARNEGIE INSTITUTION OF WASHINGTON. 



Conductivity of Vapors at High Temperatures. 



The older electric furnace has been used for a series of experiments on the 

 conductivity of vapors of compounds. The chemical features of the problem 

 were planned by Professor A. A. Noyes, and the investigation has been carried 

 out by Mr. King, with the assistance of Messrs. Weymouth, Badger, and 

 Smith. The possibility of the use of high temperatures with the furnace, 

 and of close control of temperature and other conditions, makes possible an 

 extension of work which has been done at lower temperatures with flame 

 vapors. The salt to be tested is vaporized in an extension of the furnace- 

 tube, and is then mixed with nitrogen which flows at a known rate, thus carry- 

 ing the vapor into the highly heated portion where the resistance is measured 

 between the wall of the tube and an exploring rod supported axially. The 

 concentration of the vapor may be changed by known amounts during the 

 experiment. The relative conductivity under different conditions is thus 

 measured directly, and absolute values may be obtained by checking with a 

 liquid electrolyte used in the tubes when cold. The work thus far with the 

 chlorides of calcium, sodium, and potassium has involved many variations of 

 the experimental arrangements in the course of the development of an ac- 

 curate method, and the series of final measurements has only just begun. 

 The variation of conductivity with the square root of the concentration ap- 

 pears to hold for temperatures as high as 1600° C. A satisfactory degree of 

 consistency for the readings has been obtained, and the use of temperatures 

 as high as 2000° seems to be quite within the range of the apparatus. 



Miscellaneous. 



In addition to a large number of spectrograms of iron, made with the 

 electric furnace under various conditions to check and amplify the previous 

 material, more or less extensive examinations have been made of the spectra 

 of titanium, rubidium, calcium, barium, copper, and zinc, largely in con- 

 nection with the ionization phenomena and the behavior of enhanced lines. 

 Banded spectra, especially of calcium and copper, have also received atten- 

 tion. Photographs of the cyanogen bands with the furnace have been supplied 

 to Dr. R. T. Birge and have been used by him in a discussion of Heurlinger's 

 theory of band structure and as a means of estimating temperature by the 

 location of the maximum in a band series. 



ELECTRICALLY EXPLODED WIRES. 

 The work on electrically exploded wires has been continued by Mr. Ander- 

 son, assisted by Mr. Sinclair Smith, along four different lines. 



Spectra. 

 Spectrograms have been made using wires of copper, silver, gold, mag- 

 nesium, zinc, cadmium, aluminum, tin, lead, and tungsten, in addition to 

 those of iron and nickel usually employed. In the earlier work it was observed 

 that the spectrum of copper was much weaker than that of iron or nickel. 

 The reason for this became apparent when it was found that wires of pure 

 copper, silver, or gold do not explode properly when confined in a slot in a 

 block of wood. The current passes through the wire until it is melted and 

 vaporized, then it suddenly passes between the electrodes outside the wooden 

 block, even if these are separated by a distance much greater than the length 

 of the wire. The result is a very feeble effect in the slot and a brilliant spark 



