200 REPORTS ON INVESTIGATIONS AND PROJECTS. 



ments considerably larger than are given by the arc at equal pressures, 

 though, on account of the great difficulty in obtaining measurements of 

 high accuracy in this class of work, more results from both arc and fur- 

 nace are needed to establish the details of the difference. 



The study of spectra given by the electric furnace at various temperatures 

 has been continued, material being now on hand for a number of elements 

 important in solar and stellar spectra. The range of spectrum photographed 

 is from A 2500 to A 7500. The spectra to the red of A 3900 are on large-scale 

 plates and the results will be published in a series of papers, beginning with 

 the spectrum of iron. The ultra-violet photographs thus far made are on 

 a small scale and will be supplemented when a spectrograph of higher dis- 

 persion for this region is available. The leading features in this investi- 

 gation are as follows : 



( 1 ) A number of lines are found for which low temperatures are es- 

 pecially favorable, some lines of this class being very faint in the electric 

 arc. 



(2) Other lines given by the furnace are classified according to the tem- 

 perature at which they appear, the rate of growth in intensity being ob- 

 served for increasing furnace temperatures and for the change from fur- 

 nace to arc. Among such lines pairs are selected in various parts of the 

 spectrum whose relative intensities in a given source of light where thermal 

 radiation predominates are indicative of its temperature. 



(3) Lines given by the radiation near the poles of the arc and usually 

 appearing strongly in the spark (polar or enhanced lines) are absent or 

 faint in the furnace spectrum, indicating that the electric discharge or some 

 equivalent source of high energy (probably high temperature in celestial 

 bodies) is required for the production of these lines. 



(4) In harmony with earlier work, it is found that in the arc between 

 iron terminals, the growth in intensity of spectrum lines observed in passing 

 from the outer envelope to the core of the arc corresponds closely to the 

 changes observed with increasing furnace temperatures, thus showing that 

 the relative strength of some lines in the flame of the arc is due to reduced 

 temperature rather than to chemical action. As the iron arc is easy of pro- 

 duction and study, observers not having an electric furnace at disposal may 

 employ the arc to obtain gradation in the intensity of iron lines which are 

 qualitatively in agreement with those given by various temperatures of the 

 furnace. The quantitative correspondence between a measured temperature 

 and the resulting spectrum can be obtained only with the furnace. 



(5) Striking differences appear between different elements as to the 

 relative richness of the furnace and arc spectra. Thus the furnace spectra 

 of titanium and vanadium show a larger proportion of arc lines than does 

 that of iron, though the former elements have the higher melting-points. 



(6) By the use of a small-scale spectrograph, the variations in radiation 

 with wave-length throughout the spectrum has been observed for different 

 temperatures. 



