8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 8/ 



"sunburn " or erythema (4). The percentage of the total radiation 

 which falls in this region is shown in Table 3. 



Measurements were made on the more prominent lines in the 

 spectrum of a new arc. These are shown in Table 2, column 5. The 

 differences between these values and those in column 4 are not large 

 but are greater than experimental error and probably indicate the 

 variation to be expected. 



Observations were also made upon two new arcs operating at low 

 intensity. These two arcs are to be used only as standards, and since 

 there is very little evaporation of the tungsten cathode or deterioration 

 of the quartz at this temperature and wattage, they have never been 

 run at any condition other than that of low intensity. Their output 

 is dift'erent (by an amount greater than experimental error) from 

 that of the 400-hour-old arc only in the 2,536 A. region. In this 

 region one arc is 7.6 per cent lower and the other 2.7 per cent higher 

 than the value given in Table 2, column 6. 



Figure i is a map of the high-intensity spectrum obtained with the 

 double monochromator. The eff'ective slit width used is 2 A. at 

 A 2,300 A. which increased to 12 A. at A 4,000 A. The intensities 

 are mapped above a spectrogram (taken with a type E2 Hilger quartz 

 spectrograph) of the arc operating at high intensity. The ordinates 

 in Figure i are the values given in Table 2, column 4, and are for 

 the 20-mm midsection of the arc. In order to obtain approximate 

 values for the full length of the arc exposed, multiply these mid- 

 section values by 7.1. This factor is obtained from the filter measure- 

 ments given in Table 3. 



Figure 2 is a map of the low-intensity spectrum made from Table 

 2, column 6. It is plotted above a low-intensity spectrogram of the 

 same source. The same factor, 7.1, will convert these ordinates 

 and the values in Table 2, column 6, to approximate total arc values. 



Figure 3 shows two microphotometer curves of spectrograms of 

 the 2,536 A. region. The upper curve is from a high-intensity spectro- 

 gram and the lower curve from one of low intensity. The exposure 

 times were 20 seconds for the low intensity and i second for the 

 high intensity. For the sake of comparison the slit width used in 

 the thermocouple observations is indicated. The resonance line, 

 A 2,536.5 A., which is partially self reversed at low-intensity con- 

 ditions and completely reversed at high intensity, the line A 2,534.8 A. 

 and a background of continuous (and band?) spectrum are super- 

 imposed when observed with this 5 A. slit width. The reversal of 

 the resonance line and the appearance of the background near it are 



