136 Professor Alexander Dickson on the 



unlikely to have become fixed and aceunaulated by natural 

 selection. To enter furtlier on this subject, however, 

 would lead me beyond the scope of the present paper ; and 

 I would leave the matter in the hands of those who may 

 occupy less of the position of an agnostic towards the 

 doctrine of evolution, or who may entertain greater hopes 

 that the question of the "origin of species" can ever be 

 settled, than I do.* 



5. In the foliage-leaf of Euscus androgynus, the elements 



* It may not, however, be out of place for me here briefly to refer to the 

 case of heterophyllous aquatics, and to the very interesting observations and 

 experiments of Hildebrand on certain aquatic and amphibious plants, as having 

 an important bearing on this question of stomatic distribution. 



As regards heterophyllous aquatics, it is well known that in some {e.g. , certain 

 Batrachian Ranunculi, and Cahomha) there are two forms of leaves : the sub- 

 merged, which are destitute of stomata, and the floating, which have stomata 

 only on their upper surface. In others, again, (e.g., Hii^puris) there are two 

 forms of leaves, the submerged and the aerial. While in a third category there 

 is a plant, Sagittaria sagittifolia, in which Hildebrand {Bot. Zeitung, 1870, 

 p. 17) has pointed out that all three forms may be produced ; the leaves formed 

 at the beginning of the season being submerged, those formed a little later 

 being floating ones, while the last developed are of the ordinary aerial type. In 

 connection with the floating leaves, he further made the interesting observation 

 that while the earlier developed ones are almost quite destitute of stomata on 

 ohe under side, those developed later exhibit a distinct approximation in 

 structure to the aerial leaves, the stomata on the under side being almost as 

 numerous as they are in these leaves. 



Of perhaps still greater interest are the so-called amphibious plants, such as 

 Polygonum amphibiuvi and Marsilea, which may appear in two forms: the one 

 terrestrial, with aerial leaves, and the other aquatic, with floating ones. In 

 connection with these amphibious plants Hildebrand has made some important 

 observations {loc. ciL, p. 1). In Marsilea, he made the discovery of floating 

 leaves in a plant of M. quadrifoUa which happened to be growing from the 

 bottom of a tank. These floating leaves had the stomata exclusively on the 

 upper leaf-surface, while in the ordinary aerial leaves they are nearly equally 

 distributed over both surfaces. He further experimented with other species, — 

 M. data and M. pubeseens, — and found that these also, when planted under 

 water, produced floating leaves. In the case of Polygonum amphihluin, he 

 took plants of the terrestrial form, with preponderating development of stomata 

 on the under leaf-surface, and sunk them in a tank in 3 feet of water, with the 

 result that tlie growth of the aerial shoots was arrested, their leaves decaying 

 away, while from the rhizome other shoots were produced, which in a few 

 weeks reached the surface, and spread themselves out with their leaves, now 

 developed as floating ones. 



Such facts are certainly very striking. The transition forms between 

 the floating and the aerial leaves in Sagittaria might fairly be used 

 in illustration of the steps of a supposed evolution of the one leaf- type 

 from the other — the floating from the aerial, or vice versd. The case of the 

 amphil>ious Marsileas and Polygonum, however, seems a very extraordinary 

 one ; inasmuch as here we have the sudden production of floating leaves 

 accompanying a change in the external conditions : so extraordinary, indeed, 

 as to have led Hildebrand to hazard the conjecture that in such cases 

 the aquatic form was the original one ; that the terrestrial form was slowly 

 evolved under natural selection ; and that the sudden production of floating 

 leaves de)>ends on the retention by these plants of a capacity for reversion, 

 under suitable conditions, to the ancestral type. 



