270 Experiments made by the late Dr. Henry Draper 
built above his stable in the rear of his house and which was 
completed in 1880) it was found practically impossible to use - 
carbon bisulphide prisms in this room owing to the rapid vari- 
ations of the temperature there. No definition whatever could 
be obtained with the same prisms which had performed so well 
in the main house. In consequence, the use of these prisms 
had to be abandoned and a series of experiments was made to 
obtain the spectrum by other means. First a Rutherfurd sil- 
vered glass grating of 8,640 lines to the inch was employe 
and subsequently a train of six flint glass prisms made b 
Steinheil and loaned by Mr. Rutherfurd. With reference to 
these experiments, Dr. Draper says: ‘‘ The exposures required 
v 
when using the silvered grating were so long that experimenta- 
tion was very teious; but when in addition, the definition 
did not equal that of the two bisulphide prisms formerly used, 
a change became necessary. If we could overcome the eflect 
of temperature on bisulphide it would doubtless serve our pur- 
pose best, because it is more transparent, less colored and loses 
less light by reflection than glass prisms since there are only 
two prisms needed to do the work of four flints. But the in- 
stability of a bisulphide train is so marked in the new labora- 
tory on account of the fluctuations of temperature, that we have 
not been able to depend upon it. Possibly if the prisms were 
enclosed in cotton batting or immersed in water these difficul- 
ties might be overcome.” 
With the flint prisms, the definition of the sun spectrum was 
excellent, but it was found impossible to get the line H on the 
photographic plate, through the train, The silvered glass 
the spectrum of the carbon arc, the dispersion given by the 
train between G and H was double that given by the silvered 
