338 CARNEGIE INSTITUTION OF WASHINGTON. 



which metalUc masses are composed, with corresponding decrease in 

 the number of crystals. The results are not yet ready for sunamarizing. 



I have studied the erosion of great guns at the request of the Naval 

 Consulting Board. The work thus far has consisted chiefly in an 

 analysis of the conditions causing erosion, as disclosed by microscopic 

 examination of eroded gun-tubes. I infer that materials which resist 

 erosion should combine with a high melting-point the property of 

 "white-hardness" and great thermal conductivity, density, and specific 

 heat. I have found that the collective width of the cracks formed in 

 eroded gun tubes is far more than can be explained by simple expansion 

 and contraction, and I refer this excess to cumulative cracking, caused 

 by the stopping of the mouth of each crack with a mouthful of copper 

 which its hardened steel lips bite off from the copper driving-band of 

 the projectile as it passes. 



An exhaustive series of experiments into the causes of the effects of 

 heat treatment on the properties of steel has been projected and begun. 

 In carrying out this work I shall have the collaboration of Sir Robert 

 A. Hadfield, of Sheffield, England, whose experimental laboratory is 

 well equipped for making the needed mechanical tests. The United 

 States Bureau of Standards also has indicated a willingness to collabo- 

 rate in this work. 



The output of work for the current year has been interfered with 

 very greatly by the labor of installing a new metallographic laboratory 

 at Bedford Hills, New York, and the transfer to it of the work for- 

 merly done at Columbia University. The freedom of the new labora- 

 tory from interruption, and its being exclusively for my own scientific 

 investigations, should increase the output of work hereafter. 



Nichols, E. L., Cornell University, Ithaca, New York. Systematic study of 

 the properties of matter through a wide range of temperatures. (For pre- 

 vious reports see Year Books Nos. 4-15.) 



New Absokption Bands. 



Complete maps of the fluorescence and absorption of the double 

 chlorides have been made during the year and the wave-lengths, fre- 

 quencies, and intervals for all observed bands have been tabulated and 

 discussed. This work is supplementary to the detailed study of the 

 polarized spectra of the four chlorides mentioned in this report for 

 1916 (Year Book No. 15, p. 382). 



In the course of the investigation the range of the absorption spec- 

 trum was greatly extended towards the red. The newly discovered 

 bands are discernible only when fight is transmitted through very 

 thick layers and when special precautions are taken to prevent fluores- 

 cence. They correspond in position with the fluorescence bands of 

 that region and have therefore the frequency interval of the fluores- 

 cence series instead of the shorter interval of the absorption system. 

 By means of these observations, which were made by Dr. Howes, we 



