Lcioinsoii-Lcssi/ig's Classificat'uvi of Rocks. 353 
the quantities of the oxides are not continuous functions of the 
quantity of SiO^. But if we neglect the zigzags and consider 
only the general direction (i. e. that of a right line joining the 
commencement and the end of each curve) it can be seen that: 
— the general quantity of the oxides diminishes with the gen- 
eral content of silica; the diminution of RO is rapid; the in- 
crease of R^O is much more feeble. The line R^O^i is nearly 
indififerent; it lowers slightly. If we consider the real lines in 
zigzag, we see that on these lines the zigzags corresponding 
to the maxima and the minima of the oxides (RO, R^O^, R-;0. 
RjO + RO) are repeated periodically; at all the o.i of SiO= 
there isa maximum for RO and for R^O +R0, and a mini- 
mum for R20and R'X,)3. This general periodicity does not 
take place with the acid rocks, but eiigros one can say that the 
(juantiiies of the different oxides are periodic functions of the 
content of Si( )-. At the same time it is noticeable that R^O 
increasesand RO diminishes in absolute value. This perio- 
dicity is visible also in the foregoing table of classifi- 
cation. 
{Remark of tlic reviewer. I translated almost entire the part 
relating to the relations of the oxides, but althougli Lewinsfin- 
Lessing makes the qualification that it is 01 gros I avow that 
his pretended law of periodicity, based on his curves, appears 
to me so greatly en gros that I could scarcely accept it. I have 
not enough imagination to see the periodic repetition in these 
curves). 
2. The oxides RO and R2O are antagonistic; the magmas 
rich in alkalies are poor in alkaline earths and vice versa. The 
zigzags of R-O and RO go in opposite directions, except in 
certain parts of the curves, (a) between the elcolitic syenites 
and the phonolytes ; (b) between the basalts carrying nepheline 
and those with leucite; (c) between the diorytes and mcla])hyrs. 
(d) between the granites and the quart/. porpliNries. 
I'lxcept the melaphyrs, which are perhaps metamorphos- 
ed, these exceptions take place within the limits of the same 
magma, for instance the quartz porphyries are poorer in oxides 
than the granites, on account of a higher acidity, but the rela- 
tions of the oxides between them being the same the two kinds 
of rocks belong to the same chemical type of magma, h'urther, 
the metarrior])hisni of (]uartz ])orphyries distinguishes them 
