MOUNT WILSON OBSERVATORY. 255 



tion in the ultra-violet, as far as the reflecting power of speculum 

 metal permits. 



On lines free from pole effect the measured wave-lengths are in good 

 agreement with those of other recent observers, except in a few cases 

 where differences as large as 0.002 a are found. Special attention is 

 given to all such cases, and in addition the relative values of a large 

 number of lines are being separately determined under the direction 

 of Mr. St. John by means of the plane grating. It is thought that in 

 this way the inclusion of any errors, except those necessarily involved 

 in the methods of observation, may be avoided. 



WAVE-LENGTHS IN MIXED ARCS. PRESSURE EFFECT. 



Mr. Babcock has found it feasible to extend to interference obser- 

 vations the method of simultaneous exposures to two sources of hght, 

 which has proved so powerful an aid in previous work with the plane- 

 grating spectrograph. A number of very satisfactory photographs 

 have been secured, upon which occur secondary standards from the 

 iron arc, intermingled with lines from barium, strontium, calcium, or 

 cadmium, taken simultaneously from a vacuum arc through the same 

 interferometer. Two or more of these elements may often be used 

 in the same vacuum arc, and as many of their lines are very intense as 

 well as sharp, they serve as excellent standards of wave-length. It has 

 accordingly been easy to apply this method to an examination of the 

 question sometimes raised as to the constancy of arc wave-lengths in 

 mixed arcs, i.e., arcs containing more than one element. I* or example, 

 simultaneous exposures were made to the interference spectra of the 

 iron standards and to the strong red barium lines obtained from a 

 pure barium salt upon carbon terminals in a vacuum chamber. Upon 

 other photographs the combined spectra were recorded when the 

 pure barium was replaced by a mixture of barium and strontium, 

 barium and calcium, or barium and chromium, i.e., barium chromate. 

 In the case of every line so far examined, the wave-length is the same 

 when the element is used alone as when two or more elements are 

 mixed in the same arc. This method is now to be applied to other 

 mixtures and to arcs having one or both terminals of metal, although 

 until more definite and positive evidence is adduced in support of the 

 supposed variation, it hardly seems necessary to devote much more 

 time to it. It is, in fact, highly probable that pole effect in the arc 

 is at the bottom of the phenomenon. 



The method of simultaneous exposures with the interferometer also 

 lends itself to the study of the pressure effect upon the spectrum. 

 Experience has indicated the desirability of making such observ'ations 

 over a pressure range from zero to one or two atmospheres, in order 

 not to introduce too much variation in the intrinsic width and sharp- 

 ness of the spectral lines. This, with the careful eUmination of pole 



