E. C. Pickering on different forms of the Spectroscope. 301 
uniform ; it is, however, easily explained by the frequent 
variations in the position o of the prisms in the s ectroscope 
employed. It is not to be denied that the advantages of the 
method of comparison in the determination of wave-lengths 
will more fully appear when an extended spectrum shall have 
een drawn with the data yielded by a spectroscope in which, 
as in that of Mr. Rutherfurd, all the prisms are in the posi- 
tion of least deviation for each ray in succession. 
In my former paper, I have given a method of comparing a 
given spectrum directly with Kirchoff’s chart, by means of a 
particular arrangement of the scale telescope. I have. since 
found that it is simpler and better to employ a camera 
lucida attached to the eye piece of the observing teles- 
cope. The chart is placed upon the table below and serves to 
identify the lines. To find the gee ae of a single pe as for 
instance the green thallium line, upon the chart by compari- 
son, one-half of the slit of the Sottninaites must receive direct 
sunlight while the other half receives the light of the flame 
containing thallium, the camera lucida and chart being ad- 
justed in the manner already described. A filar micrometer 
may, also, be employed with great advantage whenever the 
field of view contains, at the same time, the line the wave- 
length of which is to be determined, and one or more other 
lines of which the wave-lengths are known, or which can be 
identified upon the chart by means of the camera lucida. 
Cambridge, March 6, 1868. 
Art. XXXII—On the Comparative Efficiency of different 
nee of the cage 12 by Epwarp C, PICKERING. 
vha aon to pro- 
duce the seine di oo r ed: with the least loss of light. We 
have then to consider : ie dispersion, the loss by as and 
that by absorption. # 
1. Dispersion.—The dispersion 
of any part of the spectrum, is 
proportional to the mre inter- 
val be 
