175 



the intervening forms between E. miiyydalina and E. Risdoni. Has large domed 

 fruits and coriaceous leaves, which are often glaucous along the edges, giving them 

 an unusual hoar-frost appearance. 



Lanceolate leaves are common on the tops of branches of E. Risdoni, it being 

 not an uncommon occurrence to find the sessile, almost cordate, leaves and the 

 lanceolate leaves on the same branch. This was first drawn attention to by Bentham 

 (B.Pl. iii, 203). Fig. 1, plate 32, shows three kinds of leaves taken by me from 

 the same tree, including leaves typical of E. Risdoni and its var. elata. Mr. Deane 

 and I made a similar observation in regard to E. pulceriilenta, leaves of the two 

 shapes being found on the same twig. (See Proc. Linn. Soc, N.S. W., 1900, p. 110.) 



5. E. radiata, Hook., f., non Sieb., var. l-4i jxirtim (Fl. Tas., i, 137). 



Hooker's observations form an excellent example of the difficulties (especially 

 great in the case of the older workers) in dealing wdth plants of the mnygdalina 

 group. In making iip sets for distribution, the so-called varieties were not absolutelv 

 identical, being, in some cases, taken from different trees. 



None of the specimens are the true E. radiata, Sieb., which I have not yet 

 seen from Tasmania. Hooker {loc. cit.) attempted to define, and distributed, five 

 varieties of E. radiata, Sieb., from Tasmania. \av. 5 is var. nitida of amygdalina, 

 as I have already .shown. I will show that vars. 1-4 are for the most part E. Risdoni, 

 var. elata. Bentham (B.Fl. iii, 203) said, " one of the varieties is E. Risdoni, var. 

 elata." The whole of Hooker's types of these forms have recently been placed at 

 my disposal in quantity, so that I am able to speak with a tone of certainty I 

 previously could not adopt. Hooker was doubtful himself of these forms. He 

 said {loc. cit.) : — 



A very common plaiiL ; as above chai-acterised, it may pei'haps include several species, and, amonsf.st 

 others, varieties of ami/gdalina. The forms I have enumerated are not all of them varieties in the correct 

 sense of the term, but only states of one or more varieties, and, in some cases, of one individual, but it is 

 quite impossible to unravel them. A small or lofty tree, with a straight trunk, sometimes with a smooth 

 and sometimes a flaking or almost stringy bark. Branches more or less weeping. Leaves usually shining, 

 rather small, seldom .3 inches long, with one midrib, and very inconspicuous lateral veins, or none ; narrow, 

 sometimes very much so, though not so narr-ow as E. amygdalina usually has them. Flowers and capsules 

 always pedicelled ; the latter turbinate or obconic, rather large. 



Following is Hooker's definition of E. radiata : — 



Arbor niediocris, ramulis gi'acilibus siejjc pciidulis, foliis angusto ol!i{>ticls laiiccolatisve mediocribus 

 vix nitidis l-nerviis rcctis falcatisve, pedunculis subeUmgatis multifloris, floribus jiedicellatis, calycc obconico 

 v. clavato, operculo brcvi, capsula pedicellata. Variat insigniter : — 



1. Foliis lineari-elongatis, fructibus latioribus quam longis subturbiiialis, ore dilatato j)lano. Ad 

 E. amygdalinam tendens (Gunn : 1,07.3, 1,077, 1,102). 



1,07<^ is from "river-side, Risdon," and is E. amygdalina, Labill., aff. var. 

 tiitida, Benth. 



1,077 is from " top of Grass-tree Hill," and is E. Risdoni, var. elata. 



