TRANSACTIONS OF SECTION K. 807 



tions, it seems only likely that others may ; and consequently, that the study of 

 adaptation must enter upon a new phase in connection with the study of com- 

 parative and experimental morphology and of variation. 



Another question raised is the insecurity of our conceptions of genera and 

 other taxonomic groups, and the need of some more quantitative and phylogenotic 

 basis. We divide the Podostemacere, for example, into genera and sub-order.s 

 largely upon the degree of dorsiventrality displayed in the Howers and fruits, bu'j 

 we do not know the real quantitative or phylogeuetic value of the distinctions. It 

 also follows from the above conclusion that it is highly probable that many genera, 

 species, sub-orders, or even larger groups, may be polyphyletic. We can easily 

 imagine a group of allied or similar species, for example, all becoming more 

 dorsiventral in their vegetative systems, and at last all of them showing the efi'ect 

 in the floral organs. As the effects in the latter seem to follow very definite rules, 

 all these species may presently form a polyphyletic genus, there being no ancestor 

 which has the generic characters. The same reasoning applies to higher groups, 

 and in the very largest groups we are already beginning to perceive clearly that 

 many, if not most, are more or less polyphyletic. It is evident that this view, if 

 it prove true even in part, will help in clearing up the darkness which surrounds 

 many of the problems of evolution, geographical distribution, &c. 



8. Foliar Periodicity/ in Ceylon. By Herbert Wright. 



The high temperature and humidity of the air in most parts of Ceylon allow 

 almost continuous growth of the arborescent vegetation. There are, however, 

 nearly two hundred species which become leafless at different times of the year. 



External and internal factors affect the phenomena of defoliation and foliar 

 renewal. The climatic effect is obvious from the fact that the majority of our 

 deciduous species become leafless during our hottest and driest months. The 

 deciduous trees respond only to one hot dry period of three or four months, and 

 not to the dry part of each monsoon. Some species undergo complete defoliation 

 twice per year ; others exhibit incremental foliar activity several times per year, 

 in addition toa complete annual renewal, and many introduced species show great 

 variation during their phase of acclimatisation. In the northern districts, where 

 the rains of the south-west monsoon are very feeble as compared with those at 

 Peradeniya, the defoliation is considerably delayed. The climatic conditions in 

 Ceylon are not equable enough to allow continued development along personal 

 lines, and botanists desiring to study the personal equation in plant life should 

 select a more equable area. 



Internal forces are, however, obviously at work, as evidenced by the following: — 



(a) Species retain full possession of their foliage or put forth new leaves when 

 the temperature and dryness of the air is at the maximum. 



(b) Some species drop their leaves and remain bare during wet, cool months, 

 when transpiration is at the minimum. 



(c) Plants of the same species, on the same plots, are deciduous at periods 

 vaiying by many weeks and months. 



(d) The same species may undero'o defoliation at approximately the same time 

 of the year, though under the dissimilar climates of Peradeniya, Colombo, and 

 Mannar. 



The iiTegularity of foliar periodicity is very pronounced. There is not a month 

 when all the trees are in full leaf. 



The foliar periodicity of the evergreens is as complex as that of the deciduous 

 trees, the foliar renewal taking place annually, bi-annually, or weekly, all being 

 subject to individual variations. 



Complete defoliation and foliar renewal in temperate and tropical zones often 

 results in the differentiation of rings of growth in the secondary xylem. The 

 variation in our tropical species is so great that an exact knowledge of foliar 

 activity must be at hand before the time-value of the rings of growth can be 



