806 REPORT— 1902. 



The lowest temperature at wliicli assimilation could be detected was — 6°C. 

 This is the first well-established case of assimilation below 0° C. 



For temperatures between — 6°C. and 33° 0. it was found that assimilation is 

 affected in exactly the same way as is respiration. Provided the illumination is 

 sufficient, the assimilation increases with the temperature. At any given tem- 

 perature the leaf is only capable of a limited absolute amount of assimilation, and 

 increase of illumination beyond the amount requisite for this maximal amount 

 produces no further effect at all. A greater assimilation can only be obtained by 

 increasing' the temperature. Thus the fuudamental condition regulating the 

 assimilation is the temperature, the intensity of the light occupying a secondary 

 position similar to that of the percentage of carbon dioxide. This simple concep- 

 tion of the relation between temperatures and assimilation stands quite apart from 

 all previous views. 



For temperatures above 33° C. the result is complicated by the injurious effects 

 of the temperature. The fatal temperature for Cherry Laurel leaves is in the 

 region of 41-45° C, but the specific resistance of the individual leaves is very 

 variable. Death is shown by a rapid decrease in the respiration, but it may be 

 several hours before this ceases entirely. Exposure to light has a most marked 

 effect in increasing the resistance of leaves to the eftect of high temperatures. 



Most interesting results were obtaiued from the prolonged exposure of leaves 

 to high temperatures. It was found that the respiration of a leaf in the dark falls 

 off much more rapidly than that of a leaf in the light, and the former can in no 

 case be taken as a measure of the latter. At first assimilation and respiration are 

 equally and similarly affected by the temperature, but later the assimilation 

 ceases, while respiration is still active. 



7. On the Dorsiventrality of the Podostemacefe, ivith reference to current 

 vieivs on Evolution. By John C. Willis. 



The paper read is an extract from a forthcoming paper in the annals of the 

 E-oyal Botanic Gardens, Peradeniya, upon the Morphology and Ecology of the 

 Podostemaceffi. This order shows a very far-reaching dorsiventrality of structure, 

 both vegetative and floral. Two lines of argument are followed— one morpho- 

 logical, the other ecological. Commencing with the less modified types found in 

 the order, which show a slight amount of dorsiventrality in the vegetative system 

 and none in the floral, a series is traced tlirough the more modified types, showing 

 the progressive increase in dorsiventrality of the vegetative system followed 

 throughout by an increase in that of the floral, showing first in the spathe and 

 bracts, then in the androeceum, next in the gynaeceum, and lastly in the interior 

 of the ovary. In the next place, the same series, regarded ecologically, shows that 

 though the flowers are steadily more and more zygomorphic — a condition usually 

 regarded as an adaptation to insect visits and accompanied by a horizontal position 

 of the open flower — we have hero flowers which stand stiffly erect, and are move 

 and more anemophilous and autogamous. 



The most reasonable explanation of these facts seems to be that the dorsiven- 

 trality of the flowers has been forced upon them, without reference to any advan- 

 tage or disadvantage in the performance of their special functions, by the steadily 

 increasing dorsiventrality of the vegetative system, the latter being due to the 

 general effect of the total conditions of life acting on the hereditary peculiarities of 

 the ancestral forms, whether directly or indirectly. Now the dorsiventrality of 

 the floral organs is a character of high taxonomic value, and npon the various 

 degrees of it the grouping of the Podostemacese is chiefly founded, while it is 

 always regarded as important in other families. The conclusion drawn is supported 

 by the facts of dorsiventrality in other families, and if admitted as probable opens 

 up a number of new points of view, and raises questions which must be settled 

 one way or the other. 



If one character of importance may thus be forced upon an organ or organs 

 without reference to any advantage to that organ in the performance of its func- 



