Guanidine and Aluminium. 
473 
This agrees satisfactorily with Experiment I. and with the 
calculated percentage. 
As regards the preparation of the substance, the best results 
were obtained by the following method : — 
Approximately equivalent quantities of guanidine carbonate 
and crystallised aluminium sulphate are dissolved, separately, in 
water. Dilute sulphuric acid is added drop by drop to the 
guanidine carbonate solution until no further effervescence occurs. 
The two solutions are then mixed, filtered if necessary, concen- 
trated by boiling if the solution is not sufficiently strong, and left 
to evaporate in a vacuum desiccator. Large, well-developed 
crystals may thus be readily obtained. 
A crystallographic examination of the substance was made in 
the Mineralogical Department of the University under the direc- 
tion of Mr A. Hutchinson, to whom I am indebted for the following 
report : 
“ The crystals are short hexagonal prisms terminated by basal 
planes. They are of considerable size, measuring up to 10 mm. 
across the base, and up to 5 mm. along the axis. They exhibit 
a perfect basal cleavage. No faces other than those of the prism 
and base were observed, but on immersion for a few moments in 
water, beautiful etched figures were developed on the basal planes. 
Under the microscope these are seen to be bounded by very flat 
three-sided pyramids on an equilateral triangular base, one edge 
of which is parallel to a prism edge. The triangular markings 
on the two parallel basal planes point opposite ways and the 
planes bounding them are parallel. From this it may be con- 
cluded that the crystals probably belong to the scalenohedral 
class of the rhombohedral system. 
Sp. gr. 1'806 at 13£° compared with water at 4°. 
Examined in convergent polarised light, the characteristic 
uniaxial interference figure is seen through the base. Bire- 
fringence, fairly strong, negative. The indices of refraction for 
sodium light, determined by the minimum deviation method, 
using two natural prism faces, are as follows : 
li 0 = 1-5423, /x e = 1-4540.” 
I hope at a later date to attempt a further examination of the 
formation of double salts of guanidine and its derivatives, and in 
particular to try whether they will not, under suitable conditions, 
form vitriols and alums. 
Guanidine sulphate crystallises with half a molecule of water 
(Matignon, Compt. Rend. 1892, 114, 1432), and aluminium sul- 
phate ordinarily with eighteen molecules. Thus the total amount 
of water of crystallisation contained in the two sulphates, sepa- 
rately, is greater than that contained in the double sulphate here 
