Measuring and Becording Spectrum Bands. By T. Palmer. 281 
millimeter, which is nearly the extreme limit of any reliable exacti- 
tude. Table No. 1 shows the equivalent value of the milled head 
divisions in their relation to those of the micrometer, and which 
must be added to the primary reading thus : 
Band situated between 22 • 7 and 8 of the micrometer, milled 
head records when turned to get the centre 5, then 
22-7 + 0-025 = 22-725; 
these milled head divisions may be again divided to a ^, ^, i, ^, or 
•yV, in which case the divisions of Table No. 1 would have to be 
divided by the fraction employed. In this paper I have avoided 
entering upon any very practical results, as I wish it to be exclu- 
sively one on a new style of measurement ; however, I will, with ^ 
your permission, say in as few words as possible, which bands I 
have taken, as those to which reference may be made. They are as 
follows, and comprise the principal lines of the solar spectrum : 
See Plate CLXI., Fig. 1. 
a Solar 
B „ 
C „ 
D 
E „ 
b „ 
F „ 
h „ 
Hi „ 
H2 „ 
23 
35 
718-5 
23 
6 
686-7 
23 
9 
656-2 
22 
87 
589-2 
20 
2 
526-9 
20 
48 
517-2 
19 
42 
486-1 
17 
72 
430-7 
15 
13 
410-1 
14 
42 
396-8 
14 
87 
393-3 
A. 
Angstrom. 
Mascai-t and Angstrom. 
Fraunhofer. 
Angstrom. 
Thalen, 
Angstrom. 
For the purpose, however, of establishing a standard for evening 
work, I have not used these Fraunhofer lines of the solar spectrum, 
but have taken spectra which are readily produced by simple 
means, so that when the measurement of a band is required to be 
checked that substance which is known to give a band near to the 
one in question is placed on the side stage ; the two spectra are 
therefore compared, and the distance from that known band is 
readily determined. These I call my test-objects, they are all of 
the absorption class of spectrum, with the exception of sodium, 
which I have obtained by burning a little common salt in the 
flame of a Bunsen burner. 
The test-objects employed are as follows, and I should perhaps 
add, before proceeding to the measurements of these standard 
spectra, that in all cases of observation I always use a small cap 
