. SELF-FERTILIZATION OF PLANTS. 375 
render them homogamous. That the cleistogamous blossoms of this species are degraded 
forms of the normal kind is obvious from the presence of the **lip," as well as by there 
being four and didynamous stamens. The style elongates very much, and, under the 
pressure of the closed summit, becomes bent, so that the stigma lies between the anther- 
cells, which thus pollinate it (Tab. XLIV. figs. 25a, b, c). 
BonAGINEJE.—Sir J. Lubbock (7. c. p. 131) observes of this Order:—** As Müller has 
well pointed out, there are the widest differences in the conditions of fertilisation. Pul- 
monaria officinalis is dimorphous and sterile, not only with its own pollen, but even in 
some cases with that of a different flower, unless it belong to the different form. Echium 
vulgare has lost the power of self-fertilisation, but, so far at least as we know, is fertile 
with the pollen of any other flower belonging to the species. Other species are generally 
fertilised by insects, but in their absence perform this office for themselves ; while, lastly, 
some species, such as Lithospermum arvense and Myosotis intermedia, habitually fertilise 
themselves." di 
With regard to the order of development of the floral whorls of some of the Boraginex, 
I find the pistil lags behind the others in a remarkable manner, but subsequently grows 
rapidly. This at least is the case with Borago officinalis *, the order of which is, at first, 
calyx, stamens, corolla, pistil. The corolla, however, soon overtakes the stamens, while 
the pistil remains rudimentary for a long while. It is the same with Anchusa officinalis . 
and Myosotis arvensis. With Myosotis versicolor I find there may be recognized three 
stages. In the first the pistil remains for a time very rudimentary, then it rapidly 
grows and overtops the stamens; but, thirdly, the corolla, by enlarging, lifts the anthers, 
so that the stigma now lies on a level with the middle of the anthers. Müller thus 
describes this flower (* Nature,’ x. p. 130) :— 
* The corolla, when opening, is not only still of a pale yellowish colour, like the buds of other species 
of Myosotis, but even when not yet fully developed ; the anthers and pistil are mature at the same time, 
and the stigma slightly overtops the corolla. Hence, when insects visit the flowers in this state, their 
probosces always touch the stigma sooner than the anthers, and consequently, when flying to another 
flower, always cross-fertilize it. But, by the gradual lengthening of the corolla-tube, the anthers affixed 
to its inner side are raised till they surround and fertilize the stigma, now enclosed in the corolla." 
I have drawn three stages of growth, which illustrate the relative heights of the 
stamens and pistils at successive periods (see Tab. XLIV. figs. 33 a, b, c). It is probable 
that several small-flowered gamopetalous species are self-fertilized during the fall of the 
corolla, as by that process the anthers are dragged over the stigma. Such, for et än 
probably occurs with Myosotis sp., Anchusa, &c. 
| Amsinkia angustifolia. A very small yellow-flowered plant from Kew. The order of 
x The ejection of the ‘tit of Borage is rather peculiar. As soon as the anthers have shed their pollen, the 
calyx-lobes contract, and the petal-lobes, from having been somewhat reflexed, become concave. These two actions 
combined throw off the corolla. The appendages to the stamens are outgrowths by “ enation," and are not developed 
till late. I am doubtful as to their purpose. They seem to press upon the anthers, and so form a hollow cone, with 
a circular hole at their apex. Bees stand at right angles upon the stamens when sucking the honey from the 
.  Rectariferous appendages on the corolla; the pollen thus falls on one 2m on their abdomens; and as the stigma i is 
(meli over this ume spot, intererossing is effected: 
