BY M. B. WELCH. 483 



been observed, these species contain central ducts. Among the species which 

 he predicted would fall in the same cla-ss are E. Abergiana, E. dichromophloia, 

 E. ferruginea, E. Foelscheatia. E. peltata, E. pyrophora, E. terminalis and E. 

 Watsoniana. 



Bentham (1866). on anthereal evidence, places in the Corymbosao the 

 majority of the species then described, and now found to possess ducts. The 

 only exception is E. traclnjpltloia, which, though placed in the Micrantherae, 

 was apparently classified on imperfect evidence. 



As already mentioned, those species of Eucalyptus with ducts possess a 

 mid-rib which, for the greater part of the length of the leaf, has the xylem 

 in two more or less equal zones. Species such as E. setosa, E. rmniata, E. 

 phoenicea, E. perfoliata, E. davigera and E. tesselaris, however, have a similar 

 vascular bundle in the mid-rib, but so far, in the material examined, ducts have 

 not been found. Except in tlie case of E. tesselarig, however, fresh material 

 was not available. 



Geographically these species are distributed through Western Australia, and 

 Northern Territory, extending into Northern South Australia, Queensland, and 

 New South Wales, while one species extends into Victoria. With tliis exception, 

 species with ducts do not occur in Victoria or TiLsmania. 



From an evolutionary standpoint, the presence of these ducts in certafn 

 species of the Angophoras and Eucalypts adds another argument to the theory 

 of their common ancestry put forward by Baker and Smith. In no other genera 

 of the Tribe Leptospermeae, have similar central oil ducts been observed — at 

 any rate in those species worked on — though, as shown, the occurrence of these 

 eanals ih^ purely a specific character. The closely allied genus Tristania, however, 

 has a well developed lactiferous system in at least one species. It therefore 

 seems probable that the species of Eucalyptus with ducts are the oldest members 

 of the genus, and that this character has been entirely lost in the later types, 

 which evolved for the most part smaller fruits, isobilateval leaves and a more 

 oblique leaf venation. 



Summary. 



Oil ducts, formed by the linking up of a chain of short secretory cavities, 

 corresponding in size and formation to the oil glands in the leaf, occur in the 

 stems and leaves, both normal and abnormal, of certain Eucalypts and Ango- 

 phoras. So far, they have not been found in any other genera of the Leptos- 

 permeae, and their occurrence is purely specific. 



There is no direct connection between these central duets and the leaf, 

 petiole or stem oil glands, though both structures apparently possess an oil of 

 a similar nature. They are not kinoid in character. These ducts are not con- 

 tinuous in stem and leaves, but show a break at the petiole. They are variable 

 in diameter and length, ranging from 0.3 mm. to less than 0.03 mm. in the 

 former and from over 100 mm. to less than 1 mm. in the latter dimension. 

 They apparently function as storage reservoirs. 



No evidence has been found so far to show that these ducts occur in the 

 roots and they have not been observed in the lower portions of the stems of 

 seedlings. 



The presence of these central canals in a very limited number of Eucalypts, 

 all of the Corymbosae class, indicates their primitive character, and their oc- 

 currence also in one species of Angophora, shows apparently a close phylo- 

 genetic affinity between the two genera, Angophora and Eiicnlii'ptris. 



In conclusion I should like to express my indebtedness to Mr. J. H. Maiden, 



