J.D. Hooker, Introductory Essay to the Fiora of Tasmania. 18 
servations of Carl Sprengel and others have, however, proved 
that this is not always the case, and that while Nature has appar- 
ently provided for self-fertilization, she has often insiduously 
counteracted its operation, not only by placing in flowers lures 
for insects which cross-fertilize them, but often by interposing 
insuperable obstacles to self-fertilization, in the shape of structu- 
ral impediments to the access of the pollen to, the stigma of its 
own flower.* In all these instances the double object of Nature 
be traced; for selfimpregnation (or ‘breeding in”), while 
securing identity of form in the offspring, and hence heredita 
permanence, at the same time tends to weakness of constitution, 
and hence to degeneracy and extinction: on the other hand, 
cross-impregnation, while tending to produce diversity of form 
in the offspring, and hence variation and apparent mutability, 
yet by strengthening the offspring favors longevity and apparent 
permanence of specific type. The ultimate effect of all these 
operations is of course favorable to the hypothesis that varia- 
bility is the rule, and permanence the exception, or at any rate 
only a transitory phenomenon. 
' 12. Hybridization, or cross-impregnation between species or 
very well marked varieties, again, is a. phenomenon of a ver 
different kind, however similar itemay appear in operation and 
analogous in design. Hybridizable genera are rarer that is gen- 
erally supposed, even in gardens, where they are so often operated 
upon, under circumstances the most favorable to the production 
of a hybrid, and unfavorable to self-impregnation. Hybrids 
are almost invariably barren, and their characters are not those 
of new varieties. ‘The obvious tendency of hybridization be- 
tween varieties or other very closely allied forms (in which case 
the offspring may be fertile) is not to enlarge the bounds of va- 
riation, but to contract them; and if between very different 
forms, it will only tend to confound these. That some yar Cn 
species may have their origin in hybridization cannot be mr 
but we are now dealing with phenomena on a large scale, an 
balancing the tendencies of causes uniformly acting, whose effects 
are unmistakable, and which can be traced throughout the Veg- 
dom. In ening operations the number of hy- 
* Thus, in Zobeli. , the pollen is entirely prevented by natural causes 
: a” ase its rh flower. In kidney beahs impregnation takes 
facts: see ‘Gardener’s Chronicle,’ 1858, p. 828. : ‘ 
+ A very able and careful experimenter, M. Naudin, performed a series of ex- 
periments at the Jardin des Plantes at Paris, in order to discover the duration of the 
lead 
