Miscellaneous. 351 



that, supposing (as he conceives possible) the seeds of the most 

 aberrant flower to produce plants with similar flowers, the case would 

 not even then militate against Mr. Darwin's theory, but the con- 

 trary. In the first place, let us attentively study this Begonia itself; 

 its flowers are, like those of its congeners, normally unisexual, and 

 produced in great abundance, both males and females, in the same 

 fascicles all over the plant. The female flowers are perfectly con- 

 stant in all their characters, except that they vary in having 3-4 

 cells and stigmas to the ovary and as many wings (which is not 

 without precedent in the genus). The males have usually 6-15 

 stamens in the very centre of the flower, with no trace of stigmas or 

 ovary ; and not 1 per cent, present any deviation from this condi- 

 tion. Of those that do deviate, most have 3-5 deformed stamens or 

 rudimentary ovaries in the axis of the flower, and proportionally 

 fewer perfect stamens ; and between the normal male flower and the 

 very rare instances of a regular flower with four superior carpels 

 (more or less united in the axis) and as many hypogynous stamens 

 opposite the sepals, we find flowers with every conceivable modifica- 

 tion in number, regularity, and perfection of stamens and carpels. 

 Lastly, the abnormal carpels always bear very few ovules indeed, as 

 compared with the normal ones. Now, it is very startling to be 

 asked " is it not a * saltus' for a plant at one bound to change an 

 inferior ovary and unisexual flowers into a superior ovary and bi- 

 sexual flowers ? " but there is another way of putting the question, 

 which is more accurate, however flat it may fall on the ear: viz., 

 " is it a ' saltus ' that a Begonia should produce male flowers, in a 

 very few of which the central stamens are deformed, and in others 

 are converted into more or less rudimentary or even perfect free or 

 connate pistils?" So much for the plant. My friend proceeds to 

 say that " according to Darwin's hypothesis it would have required 

 hundreds, perhaps thousands, of successive generations to have en- 

 abled natural selection to convert an inferior ovary and unisexual 

 flowers into a superior ovary and bisexual flowers." Mr. Darwin 

 will, I think, demur to this, and still more to the rash assumption 

 that, supposing any seeds of the hermaphrodite flowers of the 

 Begonia should produce plants bearing none but hermaphrodite 

 flowers, the latter would constitute even a new species amongst 

 botanists, who would infallibly detect the true nature of the sport in 

 this (as they have in similar cases), as soon as the normal state of 

 the plant were known. In the first place, we do not know how 

 many generations have elapsed since Begonia friyida commenced to 

 bear any hermaphrodite flowers, nor how many generations may 

 elapse before all traces of unisexual flowers will be obliterated in the 

 progeny of a plant now bearing only about 5 per cent, of bisexual 

 flowers ; and it is to be borne in mind not only that these ovaries 

 are incomparably the least prolific, but further, that, from being 

 hermaphrodite, they are likelv to be self-fertilized, and, according to 

 Mr. Darwin's well-established* observations, will hence give birth to 

 a less numerous and less vigorous progeny. Nor must it be for- 

 gotten that this may be the lingering type of a by-gone phase of 

 Begoniaceee when all had superior ovaries ; for that it may be the 



