282 Cross and Hillebrand—Cryolite from Colorado. 
given is impossible, but it may be well to call attention to the 
fact that Brandl was obliged to make his determinations of flu- 
orine and the metals upon quantities of 0°1106 gr. and 0°1430 
gr. weight respectively, whereas, material was not wanting for 
_ the present analyses, the determinations having been made 
upon weights of from 0°3 gr. to 0°75 gr. It nowhere appears 
that a direct test for water was made upon the material fur- 
nished by Professor Grotb. The latter, it is true, remarks (I. c. 
. 461), “‘ Ausserdem bildet sich bei letzterem (Thomsenolith) 
in den kalteren Theilen des Rohrs ein Wasserbeschlag, welcher 
beim Erhitzen reinen Pachnolithes natiirlich ausbleibt.” The 
absence of water does not, however, seem to be hereby proven, 
but simply assumed from the close approximation to 100 of 
Brandl’s results exclusive of water. Brandl himself says water 
is wanting, but does not mention if this was ascertained by 
direct experiment. The small amount of material at his dis- 
powder, and also with a deposit of water in such quantity as to 
preclude the possibility of its having been derived from but 4 
small portion of the material experimented upon. : 
Other forms of pachnolite—In some very cellular specimens 
whose walls run irregularly and seemingly without reference 
to the cleavage of the original eryolite, are crystals of pachno- 
lite of different habit. 
* P. Groth, 1. ¢., p. 461. + Ibid, p. 462. 
