22 



DESCRIPTION. 



XCVIII. E. globulus Labill. 



Investigations have shown that this is a very composite species so far as the 

 Victorian and New South Wales plants are concerned, which should never have been 

 united with the Tasmanian plant. 



For a number of years I have been puzzled with what I regarded as transit forms 

 between E. globulus and E. Maidsni, and often discussed them with my assistant, Mr. 

 W. F. Blakely, and referred to them as the E. globulus-E. Maideni puzzle. And in 

 1921-22 we critically examined the whole of the material of both species in the National 

 Herbarium as well as the fruiting specimens in the Melbourne Herbarium. The results 

 of our investigations will be seen presently, as also those of the Rev. 

 J. H. Simmonds of New Zealand. It is mainly due to the latter gentlemen's close 

 association with the cultivated forms of E. globulus in New Zealand that led to the 

 discovery that the Tasmanian and Mainland trees were distinct species. 



In Part XVIII, p. 249, of the present work, E. globulus is fully described and 

 figured at Plate 79, whilst in Part LXV, p. 218, the juvenile leaves were described, 

 which were omitted from the former part. 



In Part XVIII, Plate 79, the following figures are referable to E. globulus :— 



1. Juvenile leaf (Coll. Labillardicrc, iu Herb., Kcw). (Tasmania.) 



2a. Juvenile leaf; 26, mature leaf. Adventure Bay, Tasmania (J.H.M.). 



3. Bud. Hobart, Tasmania (L. Rodway). 



4a. Front and back view of anther ; 46, typical form of ripe fruits. Port Arthur, Tasmania (J.H.M.). 



5a. Bud; 56, fruit (R. Gunn, Flinders' Island, 1842, No. 1070). 



RANGE 



It seems to be confined to Tasmania and Flinders' Island. The Victorian and 

 New South Wales references to its range in Part XVIII, p. 251, are mainly- referable 

 to E. bicostata. 



