398 CARNEGIE INSTITUTION OF WASHINGTON. 



PHYSICS. 



Bams, Carl, Brown University, Providence, Rhode Island. Study of diffusion 

 of gases through liquids and provision for applications of the elliptic inter- 

 ferometer. (For previous reports see Year Books Nos. 4, 5, 7-13.) 



During the course of the present year Professor Barus completed 

 for the press the report referred to in the last volume, and it will be 

 printed in September 1915. One interesting result of the application of 

 the interferometer to the horizontal pendulum need only be mentioned 

 here. In applying this method to the determination of an elongation 

 due to any cause, it was found that increases of length of the order of 

 4x10"^^ of the original length should fall within the range of measure- 

 ment, so far as the method itself is concerned. 



In the furtherance of new experiments Professor Barus has con- 

 structed a linear displacement interferometer, with the parts separately 

 attached to a single wide pier and at a sufficient distance apart to 

 admit of observation thi'ough tubular vessels up to 2 meters in length. 

 The sensitiveness remains unchanged throughout. Apart from the in- 

 creased availability of the apparatus for long vessels, it offers special 

 advantages arising from the fact that the equations are simphfied when 

 the angle of incidence is zero and from the greater steadiness of the 

 apparatus. It has in fact been possible to use it continuously in spite 

 of the commotion surrounding the laboratory. 



With the aid of the new interferometer, methods worked out in the 

 earUer report were repeated under better conditions. Thus measure- 

 ments were made at some length on the refraction of air at high tem- 

 peratures, on the dispersion of air under various interesting conditions, 

 on the effect of the intense ionization on the refraction of air, etc. 

 Moreover, the method for studying the adiabatic expansion of air was 

 further developed and the specific coefficients were obtained for tubes 

 1, 2, and 4 inches in diameter. 



In all cases of displacement interferometry a difficulty is encoun- 

 tered, inasmuch as the method of measuring in terms of displacement 

 is somewhat less sensitive than the method of counting interference 

 rings or fringes. The latter course lies equally within the scope of the 

 displacement method ; but it is the distinguishing feature of the latter 

 that the eUipses may always be found and brought back to the fiducial 

 position, however sudden or extensive the phenomenon of displacement. 

 The method, in other words, is vnthout break. 



The sensitiveness of the displacement interferometer may be in- 

 creased indefinitely by equaUzing the component air-paths and glass- 

 paths respectively {i. e., reducing the path-difference to zero in each 

 case), but the elUpses at the same time enlarge and grow too diffuse 

 for use in measurement. Professor Barus has therefore sought to 

 remedy this by the use of cur\'^ed compensators, lens-shaped bodies by 



