Greville^ on the genus Aulisciis, 
41 
they are enveloped in a siliceous case. It cannot be by self- 
division, as they would then be stationary. We are accustomed 
to hear of the unsatisfactory state of certain frustules,, being- 
attributed to their probably young condition. This may 
be very convenient,, but v^hat is meant by it ? Is a frustule 
which has arrived (by some process or other) at its ordinary 
size supposed to become more perfect by successive self- 
division ? It is evident that all this uncertainty adds greatly 
to the labour of determining both species and the range of 
variation. 
In addition to all that has been said, the following excellent 
observations of Professor Asa Gray must not be omitted : 
Everywhere/^ he says, we may perceive that Nature secures 
her ends and makes her distinctions on the whole manifest 
and real, but everywhere without abrupt breaks. We need 
not wonder, therefore, that gradations between species and 
varieties should occur From the nature of the 
case, the classifications of the naturalist abruptly define 
where Nature more or less blends. Our systems are nothing, 
if not definite. They are intended to express differences, 
and perhaps some of the coarser gradations. But this evinces 
not their perfection, but their imperfection. Even the best 
of them are to the system of Nature what consecutive patches 
of the seven colours are to the rainbow. ^^"^ Among the 
Diatomace(je particularly, the maxim Natura non agit saU 
taiim applies with far greater force than among more highly 
organized vegetables, rendering the lines of specific separation 
very hard to find. So that in the very imperfect state of our 
knowledge of these microscopic forms, it would be rash in 
the extreme to dogmatise on the subject of species. What 
is therefore most required is, a more extensive acquaintance 
with the forms of diatomic life. Materials must be accumu- 
lated before they can be reduced to order ; and, as in all 
similar cases, this can only be accomplished at the risk of 
encumbering both genera and species more or less with a pro- 
visional nomenclature. This has been the inevitable history of 
every department of progressive science. Some parties, in their 
wholesome horror of doubtful species, seem disposed to assume 
that every discovery must be an old friend with a new face ; 
and to maintain that all new species or supposed new species, 
should be kept in retentis until every doubtful point in their 
history and structure is cleared up. This would be to lock 
up indefinitely a large number of interesting discoveries, and 
to retard the progress of science in this particular department, 
which can scarcely be compared with any other. It appears 
* 'Nat. Selection not inconsistent with Nat. Theology/ p. 25. 
