112 Transactions. — Botany. 



deep, cold soil of the Pelorus Valley assisted, I have no doubt, 

 in producing the results ; still, the fact remains that death 

 followed the suspension of growth without the intermediate 

 effects. From this example we also learn how very close to 

 the verge of their chmatic range species are sometimes to be 

 found. 



Among deciduous trees the time at which the leaf falls varies 

 considerably when a comparison of species is made, and slightly 

 in the case of individuals of the same species. Thus, in this 

 part of these islands the walnut, ash, Lombardy poplar, and 

 others begin to shed in April, and are quite bare early in May. 

 The weeping willow {Salix bahylonica), on the other hand, 

 retains its leaves till June — sometimes to the shortest day. 

 But the retention of its leaves by the latter species depends 

 to a certain extent on the situation in which the trees grow : 

 for instance, where the roots are in contact with running 

 water the leaves turn yellow and fall earlier than do the leaves 

 of trees occupying warm ground. We can thus see that the 

 fall of the leaf is hastened, in species capable of assuming the 

 deciduous habit, by the condition of the soil, just as death is 

 hastened in the case of the absolute evergreen species. 



As might naturally be expected, species which lose their 

 leaves early in autumn resume them late in spring, and vice 

 versa. To this general rule I have observed a marked excep- 

 tion—the lemon-scented verbena (Aloysia citrodora), which is 

 here deciduous. Though this tree never sheds its leaves 

 before the end of May, and sometimes retains them till the 

 middle of June, it is, of all my cultivated species, the last to 

 come back into leaf. From the peculiar behaviour of this 

 plant I long suspected that in a slightly milder climate it 

 w^ould become an evergreen : this conjecture I found to be 

 correct, for at The Eocks, Queen Charlotte Sound, a specimen 

 during several winters remained in leaf. The distance 

 between the Pelorus Valley and The Eocks being less than 

 twenty miles in a direct line, the climatic difference must be 

 very small ; still, it is sufficient to turn the scale from a 

 deciduous to an evergreen habit, and, as w^e shall presently 

 see, from a herbaceous to a deciduous or semi-evergreen habit. 



This last example shows how the deciduous habit may be 

 reversed as well as induced, and, though exact experiments 

 are wanting, there is good reason for believing that the process 

 of reversion, or converting deciduous into evergreen plants, 

 is possible in all cases, though the capacity for assuming the 

 deciduous habit is confined to a limited number of evergreen 

 species. From this we might conclude that the evergreen is 

 the original form, a conclusion which is strengthened by the 

 fact that certain deciduous species are evergreen when young. 



The passage of the deciduous plant into the herbaceous 



