J 12 REPORTS ON THE STATE OP SCIENCE.— 1915. 



rim has been done away with, though it sometimes recurs as an atavism, 

 e.g., in E. urmgera; the fruit is hemispherical or even broad and 

 flat ; this has led to the ovary rising higher and higher, so that in some 

 species the valves when open in the dry state are very exserted. This 

 has all tended to allow the ready and quick escape of the seeds from 

 the fruit. These points are emphasised because I have found that in 

 closely allied species, in which the fruits are -very alike, we may also 

 expect the cotyledons to be very much alike, not to exhibit marked 

 specific difference, though slight ones can often be detected. In a 

 many-species genus like that of Eucalyptus we are sure to find some 

 factors, such as fruit, bark, &c., almost identically reduplicated; in 

 such a case we certainly should consider all the factors such as buds, 

 anthers, cotyledons, leaves, essential oils, &c., in the determination of 

 the species. 



Primary Leaves. — Almost all the Eucalypts possess the two types 

 of foliage, the primary or juvenile and the secondary or adult. The 

 term ' sucker leaves ' should be entirely discarded. Study of the 

 seedlings shows that the early leaves tend almost from the first pair 

 to assume the form of the primary type; this type is always constant 

 for the species. After a period of growth varying greatly in different 

 species and even to some extent in individuals, the secondary type is 

 adopted. The peltate form of leaf is generally the true primary type in 

 the Corymbosa group. We see a persistence of this even in adult 

 leaves in E. peltata, just as we see a similar occurrence in E. cordata 

 and E. pulverulenta. The peltate leaves, as a rule, soon give place to 

 the secondary type and so the fact of their occurrence has not been 

 adequately recognised. The sessile or petiolate forms of the primary 

 leaves are always constant and hence may be of importance in differen- 

 tiating between two closely allied species ; thus the petiolate leaves of 

 E. Andrewsi distinguish it from the sessile ovate and cordate primary 

 leaves of E. dives. The presence or absence of stellate hairs on the 

 stem and primary leaves may also be of use in differentiation, as also the 

 number and distribution of the oil-glands. This is interestingly borne 

 out in the case of E. fastigala and E. regnans, which have been thought 

 to be con-specific; but the seedlings ai-e very noticeably different, that 

 of E. fasiigata having much smaller cotyledons, narrower lanceolate 

 leaves of a finer texture and almost devoid of hairs ; the oil-glands, even 

 in the early leaves, are fairly plentiful, whilst they are almost absent 

 from those of E. regnans. The cotyledons and primary leaves show 

 that E. regnans belongs really to the ' Stringybark ' group, as the 

 cotyledons are entire and reniform and the primary leaves covered with 

 stellate hairs. E. fastigata retains the stringy bark which E. regnans 

 has lost more or less. 



To sum up, Eucalypts defy all attempts to classify them according 

 to any one character. A grouping according to barks does not agree 

 with that according to essential oils, nor either of these with that 

 according to cotyledons. Yet each is helpful in its own way in guiding 

 us to a better idea of the true phylogeny of the genus. Both the 

 cotyledons and primary leaves should be considered in differentiating 

 species, as either may give valuable infonnation. In the evolutionary 



