286 Professor Mohs’s General Reflections on 
the same kind, having only a greater extent. The mineral king- 
dom contains many well-known examples of genera. Those 
who may compare the original representations of the augite 
(paratomous augite-spar) with the hornblende (hemiprismatic 
augite-spar), will find them to agree so very nearly, as to render 
it often necessary to examine certain particular characters, before 
it can be discovered to which of the two species the varieties be- 
long ; although this is a subject which it is not our purpose to 
examine in the present place. 
The circumstance of the degrees of natural-historical simila- 
rity not being equal, is not merely unprejudicial to its employ- 
ment, but has rather the effect of rendering it more general. In 
this manner the natural-historical resemblance becomes the ge- 
neral principle of classification ; that is to say, it furnishes the 
means, according to its more distant degrees, of forming repre- 
sentations still more general than those of the genus, should this 
be of any use in Natural History. Geometrical similarity is ab- 
solute, and does not admit of higher or lower degrees. Two 
triangles are either similar to each other, or they are not similar : 
we cannot say that two among a number of isosceles triangles, 
if they have not equal angles, are more similar to each other 
than to an equilateral or a scalene triangle ; or that a four-sided 
figure is more like a triangle than a pentagon, or a circle. For 
wherever there do not exist equality of angles, and equal pro- 
portions between the sides that are similarly situate, neither 
can any general similarity exist. The exactness of this idea de- 
pends upon the circumstance that geometry takes account, and 
compares the differences, of only one property, extension. Na- 
tural History, on the contrary, must reflect upon all the physi- 
cal properties of the objects considered ; and this is the reason 
why the same determinate meaning cannot be attached to it here, 
which it would have, were we permitted to confine ourselves to 
single properties. In Geometry no classification could be pro- 
duced (and it would be superfluous, however) by means of the 
idea of similarity, because this idea does not include within it- 
self a variety of different degrees ; whereas, in Natural History, 
where a classification is indispensable, the possibility of arriving 
at one, which may be consistent in all its parts, entirely depends 
upon the different degrees of natural-historical^ similarity. 
