46 
considered the presence of nitrates in deep well waters, in 
the chalk especially, as comparatively innoxious. The 
Carisbrooke well, as I have said, is in the chalky strata, and 
is 240 feet deep, and whether it contains nitrates or not, is, 
in my opinion, on the results above detailed, a most unfit 
water for potable purposes, and my reason [for bringing the 
analysis of this water before the Society was to draw atten- 
tion to the tendency often exhibited to draw the chief 
inferences of the quality of drinking waters from what may 
be called isolated reactions, without obtaining, or at all 
events without giving weight to, other indications, chemical 
and physical, which they exhibit. 
“ Note on the Identity of the Spectra obtained from the 
different Allotropic Forms of Carbon,” by Arthur Schus- 
ter, Ph.D., F.KS., and H. E. Roscoe, LL.D,, F.KS. 
Spectrum analysis serves as our most delicate test of the 
chemical constituents of a substance. Hence it appeared 
not uninteresting carefully to examine the nature of the 
spectra obtained by the combustion of natural graphite and 
of diamond in a vacuum of pure oxygen, and to compare 
the spectra thus obtained with the well known spectrum 
of carbonic oxide obtained from charcoal The preparation 
of such an oxygen-vacuum which shall yield an oxygen 
spectrum exhibiting no other lines than those of oxygen is 
a matter of considerable difficulty. The slightest trace of any 
impurity containing carbon produces the spectrum of car- 
bonic oxide. For this reason the use of caoutchouc tubing* 
and of greased stopcocks must altogether be avoided, and 
thus the experimental difficulties are considerably enhanced. 
In order to obtain a spectrum of pure oxygen entirely 
free from the lines of carbonic oxide, a necessary preliminary 
condition of our experiment, the following arrangement was 
