426 The Rev. P. Keith on the Origin of Buds. 



month of August furnished with a number of young shoots pro- 

 truding from its upper surface. These shoots were evidently 

 formed posterior to the felling of the stem ; and if the doctrine 

 was true, they were of course traceable to the centre. Accord- 

 ingly, having taken a number of transverse sections of the stem 

 of the above Willow, I found that I could, in all cases, trace the 

 path of the progress of the bud by means of the streak of paren- 

 chyma, extending from the centre, or nearly so, to the base of 

 the shoot. There were many other streaks terminating, not in 

 shoots, but in an emerging point between the alburnum and 

 bark, which point contained, no doubt, a bud that might have 

 been protruded into a shoot in some future year, if the tree had 

 been allowed to continue growing. 



There is a capability afforded no doubt in the annual protru- 

 sion of the bud into every additional layer of alburnum, that ac- 

 counts well for the ready supply and regeneration of buds which 

 almost all perennials furnish when lopped or accidentally muti- 

 lated. The fact is evidently hostile, as far as it goes, to the 

 opinions of Du Hamel and of Knight, but it does not amount 

 to a refutation of them : for that which is true of the Willow 

 may not be true of every other tree. There are some trees in 

 which no trace can be observed of the horizontal streak of 

 parenchyma, from the origin of the shoot to the centre of the 

 stem. 



At the same time, the opinions of Du Hamel and of Knight, 

 though strongly sanctioned, are not altogether indisputably 

 established by the facts which they adduce in support of them. 

 For it may be said that the result of their experiments would 

 have been the same, whether we suppose buds to originate at 

 the centre, or at the circumference. The buds had, indeed, 

 gained the circumference ; but whence they came, or by what 

 route, there is no positive evidence to show. Yet this question 



might 



