The reputed or supposed great variation of individual Eucalyptus species has arisen probably by the attempts of botanists to found species on morphological characters alone ; and Bentham, when working on this system, experienced great difficulty in finding any constant feature upon which even to establish groups, and had to discard such differences as opposite or alternate leaves, comparative length of the operculum, length of calyx tube, etc. He finally selected the shape of the anthers as a means of classification, but even this has since been found to be defective and is open to objections. Bentham evidently felt that even this was not such a natural classification as he would wish, because in his remarks, (B. Fl., iii, p. 186), he expresses a hope that a "truly natural arrangement may be founded on a knowledge of the ' Gum ' trees in a living state, upon the proposed cortical or on any other system which experience may suggest." However, we are quite in accord with him when he states, concerning his classification, that " the groups pass very gradually into each other through intermediate forms," and our results confirm this gradation of groups; but it is the individual species that shows a comparative constancy of specific characters throughout its known geographical distribution. Not only is this the case with the botanical characters, but also in their chemical constituents, a conclusion fully confirmed by the mass of evidence we are now able to submit in connection with the species here enumerated, and as the result of a research extending over a period of thirty years.

The most serious objections to Bentham's antheral system are :- 1. That of placing in the same group, and in juxtaposition, species which to those familiar with the trees in the field, are perfectly distinct from each other; and 2. That of separating under various sections trees which by bark, wood, habit, general characters, chemical properties of their oils, kinos, dyes, etc, ought to stand near to each other. For instance, in the former case, with the smooth-barked "Mountain Gum" (E. goniocalyx), the "Bundy Box" (E. elceophord), was confounded or associated. The "Apple" of Victoria, with its red timber and stringy bark, was placed with the "Apple" or "Woollybutt" of New South Wales (E. Bridgesiana), a tree with a pale-coloured timber, and a "Box" bark; numerous other examples could also be given, and similar associations will no doubt occur to any student of the Eucalypts.

The second case could be illustrated by the "Ironbark" trees, which might be grouped together, but yet stand apart; whilst with them were included the "Scribbly Gum" (E. haemastoma), "Tallow-wood" (E.microcorys), and others-trees which have quite distinctive characters and should be separated.

Probably it has been the effort to utilise this and similar systems that has led to the acceptance of the idea of great variability of Eucalyptus species, but from the results of this extensive study we consider that only in exceptional cases do such nuances exist, even under a natural classification.

Dr. Woolls (Fl. Aust., p. 217) states, under his article on the Eucalypts: -

"Whatever merit there may be in an artificial system by which museum plants can be named, .... I do not think that such an arrangement will ever meet with much favour with those who are studying the living plants."

By adhering closely to the system which we advanced in the first edition of this research, and have followed since the publication of that work eighteen years ago, it is found that the species show, comparatively, not much variation; in fact, possess such a constancy of specific characters, that is surprising in the light of previous published literature on the Eucalypts.

Statements that a species in one locality has a smooth bark, and in another locality has a "Stringybark," and sometimes an "Ironbark," are not verified by our work. For instance, it was at one time recorded that the "Scribbly Gum" (E. hamartoma, Sm.) was occasionally found as a "Stringy-bark." Investigation proved that no such variation existed, and that the tree with a "Stringybark" is quite distinct from the smooth-bark,-"Gum" (E. hcemastoma, Sm.), and has a good geographical range and a constancy of specific characters distinct from E. hæmastoma, and so it has been separated and botanically named E. Wilkinsoniana ; it is commonly known as " Small-leaved Stringybark." The red rim of the fruits appears to be the main connecting character of these species, and this feature has little discriminative value.

The reputed variability of the species of the genus might possibly have arisen, because -

1. Original descriptions were so brief, and material too indefinite, as to be practically useless. Dr. Woolls was of this opinion, for he states : ' When I find writers of some eminence referring very different trees to the same species, I cannot but see the inadequacy of the descriptions hitherto relied on." All such doubtful descriptions have been ignored in this work, as there appears to be no finality in trying to match material to such vague diagnoses and fragmentary specimens.

2. Too much reliance was placed on herbarium specimens. Botanical systematic work can generally be carried out on dried material in most instances, but in the case of the Eucalypts it is not the only evidence needed, as some of the essential natural characters are not represented in such material.

3. Common names are used too indiscriminately. In almost every work on the Eucalypts, one finds a number of common names appended to the botanical one. In the light of our present knowledge most of these common names can be shown to refer to distinct species. In this work the common names have been subordinated to the botanical, being considered of. secondary importance.

4. Sufficient attention has probably not been given to field botany in the determination of many Eucalyptus species. Dr. Woolls may be again quoted in that he states : " Trees placed by botanists under one specific name would never be so considered if studied in nature, for there the specific differences are so marked that no one would ever think the trees were one and the same species." Our researches confirm this statement; and numerous instances of this fact might be given here.