MONOCLINIC DOUBLE SULPHATES CONTAINING AMMONIUM. 
61 
are necessary in order to separate their compounds from each other. Extreme 
closeness of the crystallographic characters of their crystallised salts is, therefore, 
precisely what one would expect to find. The very opposite is the case for the 
alkali metals, which latter were chosen expressly by the author (Tutton) as belonging 
to the first group of the periodic system, exhibiting the electro-positive property at 
its maximum, while at the same time showing large differences (46 between K and 
Eb and 47 between Rb and Cs) in atomic weight. Moreover, the crystal system of 
the ethylsulphates of the rare earths is the hexagonal, and a comprehensive 
comparison of the author’s results on the rhombic (simple sulphates and selenates) 
and monoclinic (double sulphates and selenates) series has shown that the system of 
crystallisation has a preponderating effect on the angular differences, the latter being 
the smaller the higher the degree of symmetry. It might have been expected, a 
priori , that the monoclinic pair, say, K 2 Mg (S0 4 ) 2 .6H 2 0 and Cs 2 Mg (S0 4 ) 2 . 6ELO 
would give closer angular values than the more symmetric rhombic pair, for example, 
K 2 S0 4 and Cs 2 S0 4 , since there is a common “dead weight”; but experience proves 
the contrary, thus emphasising the effect of difference of system. Now the hexa¬ 
gonal is a higher system in degree of symmetry than the rhombic, and the angular 
differences (which the author found very small yet clear and unmistakeable in the 
rhombic sulphates and selenates, although not so large as those shown in the mono¬ 
clinic double salts) to be expected in the hexagonal system, other factors remaining 
the same, are still smaller, and might easily fall within the limits 2'-3' or there¬ 
abouts. Jaeger has, in fact, realised such a series, and his results are precisely what 
the author (Tutton) would have expected. It cannot be emphasised too strongly 
(a) that the author’s tables of angles of the whole 31 double sulphates, the 9 double 
selenates yet investigated, and the whole 10 simple sulphates and selenates, show a 
general correspondence of mean and calculated values within 2' for the rhombic simple 
salts and at most 5' (probably only 3') for the monoclinic double salts, and that the 
constants given by the author for each individual salt can certainly be relied on within 
these narrow limits ; and (b) that the deductions made by the author, such as the 
important one regarding the observed progression with the atomic weight of the 
alkali metal, are unquestionably valid, since the angular differences in question 
between the salts under comparison are far and away greater than these small limits 
of possible error, indeed in the maximum amounting to 2j°. 
The author has no doubt of the probable accuracy of the work of Jaeger, the 
results agreeing perfectly with the author’s expectations from the high symmetry 
displayed. The contribution is, moreover, a very welcome one at the present 
juncture. But if Jaeger’s deduction, as the author understands it, is that literally 
the axial ratio a : c of his hexagonal salts is rigorously the same for all the salts, the 
minute differences being considered only chance variations from this mean form, then 
the author must hold that this view is not a reasonable conclusion, and that it 
cannot be accepted ; it could in any case only be possibly true at some definite 
