(a) Those yielding oils consisting largely of the terpene pinene; either dextro-rotatory or laevo-rotatory.

(b) Those yielding oils containing varying amounts of pinene and cineol, but in which phellandrene is absent.

(c) Those yielding oils in which aromadendral is a characteristic constituent, and phellandrene usually absent.

(d) Those yielding oils in which the terpene phellandrene is a pronounced constituent, with piperitone mostly present.

Of course connecting species and sub-groups link up the larger groups above, and such connections are indicated throughout this work by the system of classification adopted.

The suggestion that the line of sequence is through E. saligna and E. botryoides to the "Ironbarks " was indicated, not only by the chemistry of their oils, but also by that of their kinos; the richer cineol oils also show an association with the "Ironbarks," particularly through that of E. sideroxylon, and they are all grouped in this way to illustrate the sequence.

The greater number of the Eucalypts yield oils consisting largely of cineol and pinene in varying proportions, with an absence of phelkinclrene. It thus appears that phellandrene made its appearance late in the evolutionary arrangement of the genus, and if this is so then piperitone is quite a recent formation-, even more so than the other characteristic constituent-aromadendral, which is also found in some West Australian species.

The suggestion that the "Boxes" and associated "Mallees" descended through the " Stringybarks" was derived from the oil of E. obliqua, a species with a most extensive range, the oil of which always contains aromadendral, and this characteristic constituent increased in amount as the typical "Boxes " were evolved.

A good many of the "Mallee" oils do not contain aromadendral, they being evidently more closely associated with the "Gums," and are here arranged in that manner. Such species are E. Morrisii, E. pumila, etc.

As the genus evolved the venation of the leaves and colour of the timbers changed in agreement with the alteration in the character of their oils. All the species belonging to the "Peppermints," the "Ashes," and associated species have white timbers, while in the earlier members they are usually red.

The genealogical table we now submit shows, in its general arrangement, the lines of sequence through which the Genus apparently evolved. Of course, it was not possible to depict diagrammatically, the most difficult sub-divisions which,.to any student of the Eucalypts, must become apparent, but if the table is considered broadly, the general grouping will be evident, and this arrangement is supported by botanical as well as chemical evidence.

Passing onward from E. pilularis a well-defined group of trees is reached, the oils of which contain phellandrene as a pronounced constituent, and instead of aromadendral being present, this constituent has been replaced in these oils by the peppermint ketone, piperitone; consequently the leaves when crushed give an odour of peppermint, and for this reason the trees are generally known, vernacularly, as "Peppermints" ; the more pronounced of these are E. piperita, E. dives, E. amygdalina, E. vitrea, etc. When the first Eucalyptus oil was distilled in Sydney in 1788 the leaves utilised were those of E. piperita, known locally as "Peppermint," so that the first-named species from New South Wales owes both its vernacular and specific names to the presence of this chemical constituent in the oil; and the introduction of the utilisation of chemical constituents in aiding the diagnosis of Eucalyptus trees dates as far back as the foundation of Australia. Results of recent investigations show that the value for determinative purposes of chemical constituents in the tree, as adopted by those early naturalists, was on a sound basis, and in the light of our present knowledge this determination is recognised as being of the greatest value in deciding differences between Eucalyptus trees which are morphologically closely allied.

It will be seen from plates VI, VII, and VIII that the venation of the leaves of trees giving phellandrene-bearing oils has an acute spreading arrangement inclining to run parallel to the mid-rib, which appears to be the completion of the gradual alteration of the leaf venation of the Genus, which commenced with those species closely associated with the Angophoras. Piperitone is usually associated with phellandrene in the oils of species showing this venation, although its occurrence in some of them could not be decided with certainty; but it is probable that many of the constituents found in these oils are present in traces in many of them. Pinene, also, probably runs through the whole series, of course, diminishing more and more as it is replaced by phellandrene or other terpenes. Phellandrene appears to be present in a maximum amount in the oils of E. dives, E. Andrewsi, E. radiata, etc, and these species show very clearly the characteristic venation for this group. In the lanceolate leaves of these species, too, the marginal vein has receded so far from the edge that often a second one has commenced to form. In the leaves of this group, the reticulations between the more prominent veins in the leaves belonging to the cineol-pinene group have become still more subordinate, and consequently more room is given for the formation of oil glands, and thus the yields of oil from many species of this group are large. The black dots in the photographs show the position of the innumerable oil glands in the leaves. The reproduction of the venation can be carried out very successfully by photography, the fresh leaves being used. These are printed directly upon the paper in strong sunlight, and the prints thus obtained can be reproduced by any of the well-known photographic methods.

In October, 1901, we read a paper on this subject before the Royal Society of New South Wales, and demonstrated this alteration of leaf venation in agreement with the chemical constituents, by the aid of a series of lantern slides made from the photographs taken directly from the leaves; the completeness of these can be judged from the reproduction of the leaves in the illustrations (plates I to viii).