96 
+ 
REGION 1. REGION 2. | REGION 3, 
CENTRAL VIOLET. Bur. OUTER YELLOW’ 
ELEMENTS. ee alee, 3 Lan ees |... 
ate 1. ate 2. 
Ex posure! Exposure ae 4h. ae 4. || Plate 5. ak 6. 
Beda git aan Seconds. Seconds. )||Seconds. |Seconds. 
CORES cera thas ea bese. 313 450 60 27: 1 8 
Mee certo eee en apres 84 107 29 91 47 T4 
Se LE rok tates 42 RAS Aree oe A 14 17 3 5) 8 15 
Aiea tt SEE ARR oy tae 7 sete 0 20 0 1 0 1 
PER eee oye nck Vaan, Bs ok ome ence etauns (i ate 3 4 2 3 2 3 
212M (cane, 5 al BS a ot Sel so 2 3 2 2 2 2 
Ny iT ogee oe SO Staite, 46 Mrin ea emee ae 2 2 0 2 1 1 
Oar. : 0 2 0 0 0 0 
12S SAR Sey Oe SEED oes ied Aiea, 1 2 0 1 1 al 
Winee ech eek eee ae lo cau ae 1 1 0 1 0 i 
Te setae one Meee Ger mae ta. 1 1 1 1 0 il 
IWOUAGONIINOM a. caidare se ane 4 50 0 1 0 0 
Baintiorbimnred: <i. si sweetie 29 37 38 29 6 13 
To. No. lines identified....... 421 709 97 391 62 107 
To. No, lines visible.......... 454 796 135 421 68 120 
Whole number of lines visible, 1994. 
Whole number of lines measured, 1842. 
Whole number of lines identified, 1787. 
The table shows that the lines rapidly decrease in number as the slit 
is moved from the center of the arc to the outer edge. This is due chiefly 
to the fading out of the carbon bands, only a few of the stronger carbon 
heads being visible in the outer sheath. But the fading out is not confined 
to the carbon, as the lines of all the other elements of the table show the 
same tendency, though to a much smaller degree. 
The lines of the table (which extends from 7 —3092 to 7 =45015) were not 
only measured and identified, but their intensity was estimated at three 
points—at the upper end near the positive carbon, at the center and at the 
lower end near the negative carbon. The Ca and Fe lines were relatively 
much stronger in the outer regions than the lines of any other element. 
All the lines of any one element did not fade out at the same rate. Neg- 
lecting for the present the thickening of the lines at the poles and consid- 
ering each region as a whole, it may be concluded that the differences ob- 
served in the spectra are due chiefly, if not entirely, to temperature dif- 
ferences, and that there are no lines having maxima in the outer regions. 
A study of the photographs taken under the three other conditions before 
mentioned confirms the conclusion. 
The spectrum of the inclosed electric are (a Helios Electric Co. lamp) 
did not differ from that of the open arc when the are was first started. 
But after a minute or two the are shortened and the metallic lines began 
i ee 
