380 



anthereal system. Incidentally it may be pointed out that at p. 394, I have shown 

 that venation which, in an important degree (though not exclusively) is dependent on 

 the structural requirements of the individual leaf, varies with the width of the leaf. 

 This is a contributing cause to variation in the venation of leaves belonging to the same 

 species. 



Messrs. Baker and Smith have supplemented existing evidence in regard to the 

 morphological definiteness of — 



(a) The Corymbosse (Bloodwoods), 



(b) The Renantherse, 



already strongly differentiated by taxonomists, but they leave, as indeed others do, the 

 vast intermediate group, in much the same rather indefinite position as heretofore. 



Their classification being more or less a chemical one, it is the botanist who 

 has to help the chemist towards a natural system, or rather, the classification of plants 

 lies with the botanist, and it is his duty to see if he can obtain any points from the 

 chemist to help in forming a natural one, which will eventually be, as has been pointed 

 out for very many years, based on the accumulation of data from all sources. 



This seems an appropriate place for the following passages from my pen in Proc. 



Roy. Soc. Tas., 1914, p. 24 :— 



" It is unsafe to generalise in regard to the composition of . . . oils from very few distillations 

 . . . Very many additional oils are required even for generalisations. Before a complete research can 

 be made, a full series of oil-determinations in regard to a particular species should include leaves taken 

 every month of the year, and for as many years as possible, as the meteorological conditions of any year 

 differ from those of every other year, in spite of the search after cycles by meteorologists. Leaves should 

 be collected from the lower branches and from those at the top, from those along the periphery (of the tree) 

 and from those at a distance from it. Then we require leaves from trees of various ages and sizes, from 

 trees growing in as many districts as possible, and in situations exhibiting as much accommodation to 

 environment as possible. The above refer to spontaneous trees ; the variation that takes place in cultivated 

 trees is almost a sealed book. 



Every charge of leaves submitted to distillation should be backed by specimens in the herbarium, 

 so that any questions that may arise at any time in regard to anomalies, or reputed anomalies, connected 

 with the oil results, may be considered in connection with the corresponding botanical material. The 

 referential material in regard to oil-analysis should be at least as complete as systematists find necessary 

 in their investigations of a species. As regards every oil referred to in literature, there should be a 

 schedule of particulars as to the tree which yielded it, date of collection, and so on. Systematists are by 

 no means free from blame as regards their work. Mueller has placed us under the greatest obligation in 

 regard to his pioneering monograph on Eucalyptus, a foundation on which all succeeding workers must 

 build, but in the vast majority of his plates he gives us no details as to the specimen figured. In effect, he 

 says, ' This is Eucalyptus of such and such a species, never mind whether it is the type, or a South 

 Australian or Queensland form of it ' . . . In the vast majority of plates the types certainly are not 

 drawn, and what particular form, attributed to the species, we can only guess at. The value of a botanical 

 drawing may be very greatly discounted if the precise locality, date of collection, and even the name of 

 the collector, be omitted. If these particulars cannot be given, the plate should not be published." 



As an illustration of how careful one should be to give the amplest data in regard 

 to the oil from every charge of leaves, Mr. F. W. Wakefield, botanist to an important 

 oil-distilling firm, told Mr. Blakely (one of my botanical assistants) and me that he 

 could obtain three different oils from E. radiata (Australiana), according as the trees of 

 that species grew on ridges, sides of hills, and flats in the same district. 



