172 REPORTS ON INVESTIGATIONS AND PROJECTS. 



pressure as given by Humphreys and Duffield, indicates that there is no very 

 direct relationship, since there are many lines affected in very different de- 

 grees by the two influences. A statistical treatment of a large number of 

 lines, however, shows that in general there is a fair agreement between the 

 order of magnitude of separation and displacement, pointing to the conclu- 

 sion that the physical bases of the two phenomena, if not the same, are at 

 least closely related. 



The changes in spectra with varying temperature of the radiating vapor 

 have been studied by means of the electric furnace according to the plan 

 described in the preceding report. A classification is thus made of the lines 

 of any element based on their response to change of temperature in the 

 source, and we obtain for any line which is found in the furnace-spectrum 

 not only the approximate temperature at which it appears but the rate at 

 which the line changes in intensity with increase of temperature. It is found 

 that in the case of some substances almost the entire arc spectrum can be 

 obtained at the higher temperatures of the furnace. Other elements show 

 many strong arc lines which are absent in the furnace, a fact which may be 

 explained usually by their appearance and behavior in the arc. Several minor 

 alterations in connection with the furnace have given greater efficiency and 

 better results, and the material for the study of the spectra of a number of 

 elements having many lines is nearly complete. 



Mr. Gale spent the months of March, April, and May in Pasadena engaged 

 in an investigation of the spectrum of the electric spark under pressure. 

 Reference has already been made to some of the results found in the course 

 of his work, but on account of their important applications to certain astro- 

 physical problems they may well be repeated at this point. 



(i) The enhanced lines remain bright under pressures at which the great 

 majority of the other lines are reversed. A probable explanation of the 

 intensity and persistence of the enhanced lines in the spectrum of the sun's 

 chromosphere is afforded by these observations. Since the pressures em- 

 ployed in the laboratory investigations were nearly the same as those present 

 in the sun's reversing layer, it is evident that at the sun's edge we should 

 expect the enhanced lines to appear bright when the majority of the arc lines 

 are still dark. This is in agreement with observations. Similarly the arc 

 lines which remain bright under pressure are in almost all cases prominent 

 in the spectrum of the chromosphere. 



(2) The enhanced lines are displaced more than the arc lines at the same 

 pressures. In an investigation of the spectra of the sun's limb it was found 

 that the enhanced lines are displaced more than the other lines, and a similar 

 effect is present in the spectra of the stars Sirius and Procyon. The labora- 

 tory results, accordingly, appear to furnish us with a means of determining 

 quantitatively the pressures in the atmospheres of certain stars. 



A short investigation has been made by Mr. King of the relative wave- 

 lengths of lines given by the arc and by the spark. The long-focus Littrow 

 spectrograph, giving high dispersion with freedom from astigmatism, com- 



