OBSERVATIONS FOR YOUNG BOTANISTS 165 
having swept out the pollen from the anther-tube, and is now 
ready to receive it from another flower ; but in the groundsel 
it remains included within the anther-tube and so gets at once 
dusted with the pollen. Chickweed opens its flowers and 
yields a minute drop or two of honey on warm sunny days ; 
but in the cold weather and all through the winter if there be 
no snow it goes on seeding within its closed-up flower-buds. 
Again, the flowers of the violet set no seed in this country, 
though they do so in America and Italy ; but if you turn up the 
leaves in summer you will find numerous buds which never 
open, but set an abundance of seed. They are called 
“ cleistogamous,” a word meaning “ concealed union.” 
Wind-fertilised flowers are often small and insignificant, as 
on nettles, docks, sorrel, and appear to be as easily pollinated as 
the preceding, to judge by the seed set ; for a characteristic 
feature of both is the extraordinary abundance of seed they 
bear. Chickweed and groundsel and other similar weeds will 
soon completely over-run a garden, if allowed to have their own 
way. Another peculiarity is that, if we compare the distribution 
of plants over the globe, those with large flowers which are 
regularly pollinated by insects are limited in range ; whereas 
the others are cosmopolitan wherever the climate permits of 
their existence. Thus I have before me the annual nettle 
{Urtica ureits) so troublesome in gardens, from Monte Video, and 
the sheep’s sorrel {Rumex Acetosella), from the Falkland Islands; 
chickweed from Campbell’s Island ; the stork’s-bill {Erodiian 
cicutarinm), from Patagonia, &c. 
With regard to insect-fertilised flowers, they are mostly 
very conspicuous, and, as stated, limited in range. They are, 
moreover, not nearly so fertile, as a rule, when compared with 
the preceding. Thus a particular plant of an orchid, a species 
of Dend robin III, in its native home bore 40,000 blossoms ; but 
produced only one pod. 
Many flowers are regular, i.e., have the parts of each whorl 
exactly alike, as the butter-cup. When such is the case they 
are mostly terminal on branches, and can be visited on all sides 
equally well. Many are irregular ; they are then applied more 
closely to the stem, so that an insect visits them in one w'ay only, 
or from the front, as we might say. 
These have all been descended from regular flow'ers, for they 
not infrequently “revert” to regularity, as may be sometimes 
seen in the topmost flowers of a spike of larkspur, foxglove, 
snapdragon, pelargonium, &c. ; and it is believed that the 
irregularities have arisen through the actual power of response 
to the mechanical strains to which the parts of the flower have 
been put by the weights and pressures of the insects themselves. 
Thus, e.g., the calyx-tube of the Labiates has to support that of 
the corolla, but it is far too slender, as a rule, to bear the 
enlarged upper part together with the weight of the insect 
on the “lip.” To strengthen the calyx nature has run up 
