278 SMITH -. 



phenomenon undei' observation. The impression left on the 

 writer's mind is that he was rather more strict in later months 

 and did not record certain species, whose new foliage was 

 not very sharply differentiated and which were included in 

 December. The number to be deducted for this personal 

 factor cannot be estimated, but it probably renders the results 

 for December less reliable than the later ones, , 



Closer analysis showed that there were a number of species 

 in which growth was almost continuous, and these appeared in 

 nearly all the records. 



Closer Analysis of the Phenomenon of Periodic 



Growth. 



On considering the subject more closely the phenomenon of 

 rapid periodic growth was seen to be very varied in its nature. 

 Roughly speaking, as is indicated by Wright (32) in his paper 

 on foliar periodicity, the species exhibiting this phenomenon 

 may be divided into three classes. 



In the first class is included a number of species in which 

 growth is almost continuous in point of time, taking the tree 

 as a whole. These trees, such for example as Duabanga 

 moluccana and Saraca indica, appeared in nearly all the 

 records made. In these cases the growth is not, as a rule, 

 general over the whole tree. Any particular branch axis 

 passes through a resting period before growth begins again. 

 The different branches, however, elongate at different times, 

 so that, taking the tree as a whole, there was some new growth 

 to be observed on it at all or nearly all the times of observation. 

 The amount of new growth in such trees varies, sometimes it 

 occurs in only a few twigs, sometimes in a considerable number, 

 but not, as a rule, over the whole tree. 



In the second class we may place those trees which show new 

 growth four, five, or six times in the year. Examples are 

 Eugenia Gardneri and Theobroma Cacao. As the periods of 

 new growth are fewer so the growth is correspondingly more 

 general when it does occur. Sometimes the whole tree shows 

 it, sometimes only part of it, but it is not restricted to a few 

 twigs as in trees of the first class. 



