84 PHILOSOPHICAL NOTES ON 



The female flowers have five petals, and most of the male 

 flowers only four petals. In B, Knoxdesyana I found a male 

 flower ^vith the rudiment of a fifth petal, indicating that the male 

 flowers may at one time also have had five petals.* 



Begonia metalUca has male flowers with two large and two 

 small petals. B. semperfiorens has the same feature, but B. sem- 

 perjiorens gigantea has only two petals in its male flowers. 



Here is a ver^' distinct genus which may have had at one time 

 both its male and female flowers five-petalled ; it passes into a 

 type, which now includes the majority of the species, with four 

 petals in the male flower. Then two of these are dwarfed in some 

 species, and eventually suppressed in others, so that now we have 

 also types with two petals in the male flower. 



Just the reverse of this of course may also have hapj)ened. 



We have innumerable instances of dwarfing and total suppres- 

 sion of stipules, and enlargement of them in other cases. What 

 are called stipels in leaves are only abortions of leaflets, as is clearly 

 seen in Hardenhergia Comptoniana, where leaflets are often 

 replaced by stipels, and vice versa. In other cases stipules are 

 only visible during the expansion of the bud. They drop off very 

 early. In manv cases this might be a stage prior to total 

 suppression. 



In the Cactaceai, leaves are : reduced to mere apologies of such 

 organs and drop off early, the stem enlarging and taking on the 

 leaf -function. In some genera, the leaves, even in their dwarfed 

 state, never appear. In flowering stems, leaves are dwarfed into 

 bracts, and are often as minute as stipels. 



Then we have an infinite variety of what are called teeth on 

 the margins of all descriptions of leaves, such as the teeth of rose- 

 leaves, holly-leaves, agave-leaves, aloe-leaves, &c., &c. All these 

 must have had some common descent, that is, they must have all 

 originated from homologous j^arts in some very distant ancestor. 

 The thing is to find out what that ancestor was, and what these 

 remnants could have originally meant. 



In Nicholson's Dictionary of Gardening, under the heading of 

 Teeth, the only definition I find is " any kind of small divisions." 

 Naturally these teeth, like many other parts, are sometimes totally 

 suppressed, and then we have what are called entire leaves. 



* Begonia Socotrana has normally six segments in its female flower. 



