136 NATURAL SCIENCE. Aug.. 



leaf and the ensiform one of the Iris. The former has a well-marked 

 dorsi-ventral structure and a clear differentiation into distinct sheath and 

 blade, separated by the characteristic ligule. Moreover, anatomically, 

 the blade is that of a normal aerial leaf. The Iris, on the other hand, 

 has equally well-marked isobilateral structure and properties. It is 

 dangerous to draw conclusions from similarity in shape. Is the 

 tendency towards a cordate form in the geranium leaf the result of a 

 former floating habit, and the still more divided outline in some 

 species evidence of a former submerged state ? The hypothesis of 

 the origin of the aloeidal type of leaf from an ancestral form like 

 Stmtiotes aloides seems as gratuitous. 



In the section headed " Venation," the author again attempts to 

 show that the terrestrial leaves of endogens are primarily aquatic, but 

 have taken on aerial features, such as reticulations, stomata, etc., the 

 idea being that a parallel or curvinerved venation is an aquatic 

 adaptation. 



The facility of tearing seen in the submerged leaves of Nymphaa, 

 in those of bananas, palms, and blades of wheat, is referred to an 

 enfeebled structure primarily due to a watery life. One can, however, 

 see a nearer cause in the case of leaves like the banana and palm, 

 which are not at once shed from the stem, but gradually droop. 

 This process will be aided by the tearing, and they will 

 occupy less space, and cause less drag on the stem, than if they 

 remained entire. We think, too, that a better explanation of the holes 

 in Monsteva leaves might be found than descent from an aquatic 

 fenestrated leaf like that of Ouviranda. 



Finally, as regards the flower of endogens, the author suggests 

 that the local stimulations set up by insects may equalise, if not 

 surpass, the degrading influences of an aquatic medium. Presumably, 

 this has happened in three important aquatic orders — Hydrocharidese, 

 Pontederiaceae, and Alismaceae — where the flowers are certainly con- 

 spicuous. As, however, a large number of genera of Naiadaceae, 

 Typhacese, Juncaceae, Cyperaceae, and others are greatly degraded, it 

 is claimed that " the evidence from coincidences is accumulative, 

 while comparative anatomy justifies the conclusion that there has 

 been a distinct ' cause and effect ' in the reduction of the floral 

 structures." 



The evidence given, however, does not seem particularly con- 

 clusive, comprising a derivation of the spike or spadix of Aroids, 

 Typhacese, Potamogeton, and others from the branched inflorescence 

 of Alismaceae and Juncagineae by suppression of the flower stalks 

 and perianth whorls, and the reference of the quaternary arrangement 

 of the floral organs to a reversion to a very primitive condition of an 

 opposite and decussate arrangement. The suggestion is put forward 

 in a foot-note that the use of the lodicules of grasses is to act as 

 reservoirs of water ; it is, however, not easy to see why water 

 reservoirs should be wanted in a grass-flower. 



