421 
from the Stearopten of Oil of Peppermint. 
that the formula I announced for the sulphodadylate of lime, 
was SO 3 . Ca O + Hig, and not the formula which Mr. 
Walter attributes to me. 
The substance obtained by Mr. Walter in acting on the 
stearopten with perchloride of phosphorus is of great import- 
ance, as giving some evidence of how far this essential oil acts 
as an alcohol. The action appears to be very complicated, 
and hence it is to the analysis alone that we can look for 
explanation. Taking for the best analysis the highest car- 
bon and the lowest hydrogen, the results of Mr. Walter 
Best analysis. Mean of six. 
Carbon 70*55 70*01 
Hydrogen... 9*89 10*31 
The chlorine, determined once, gave 20*87. 
The most natural formula is to suppose that, as with alcohol 
or acetone, water is eliminated and decomposed, a sort of mu- 
riatic aether being produced. Hence the formula should be 
' C 20 Hjg Cl or C 21 Hjg Cl. The numbers are, 
C -20 69*26 C 21 70*32 
Hi 9 10*72 H^g 10*36 
Cl 20*02 Cl 19*31 
100*00 
100*00 
Here with the C^q of Blanchet, there is again too much 
carbon and too little hydrogen, but with the Cgi as on my 
view, the numbers are more nearly true. But Mr. Walter 
suggests that the hydrogen may be 18 atoms in place of 19, 
and then the formulae give 
C.0 
Cl 
69*6 
c^, 
70*71 
10*3 
H,s 
9*87 
20*1 
Cl 
19*42 
100*0 
100*00 
Thus we have still on Blanchefs basis too little carbon 
and too much hydrogen in the formula. On the basis of Cg^ 
the formula becomes much more likely to be true; at all 
events the third formula suggested by Mr. Walter C^q Hjy Cl 
is quite unnecessary. I consider this body to be chloride of 
menthene. 
Although we should find similar examples in the products 
formed by the action of chlorine on the stearopten, yet 1 will 
not enter into any discussion relating to them, as from the 
analytical results, and other circumstances, it is evident that 
the final and definite products of that action have not yet 
been obtained. In Mr. Walter’s formulae there come into 
