516 



VII. INFLORESCENCE (in part). 



D.— ANDROECIUM. 

 ANTHER. 



Historical. 



Beniham, 1866. — The study of the Eucalyptus anther practically begins with 

 Bentham's "Flora Australiensis " ; at all events, he first used this organ for the 

 purpose of classification, often referred to as " Bentham's Anthereal Classification." 

 It becomes therefore necessary to ascertain what he said. 



" This usual form (of stamen) is a stalk called the filament, bearing at the top an anther divided 

 into two pouches or cells. These anther-cells are filled with pollen, consisting of minute grains, usually 

 forming a yellow dust, which, when the flower expands, is scattered from an opening in each cell. When 

 the two cells are not closely contiguous, the portion of the anther that unites them is called the connec- 

 tivum (connective). (B. Fl. I, xv.) 



". . . . Stamens numerous, in several series, free or very rarely (Eudesmiew), very shortly 

 united at the base into 4 clusters ; anthers versatile or attached at or close to the base, the cells parallel 

 and distinct or divergent and confluent at the apex, opening in longitudinal slits or rarely in terminal pores, 

 the connective often thickened into a small gland (my italics), either separating the cells or behind them 

 when they are contiguous .... (ib. Ill, 185). ... In the meantime, as far as I can gather 

 from the information supplied, it appears to me that among large trees the majority of the " Stringy-barks " 

 are to be found in my first series with reniform anthers, and of the. " Iron-barks " and " Box-trees " in 

 the following three series with very small globular or truncate anthers, that other marked peculiarities in 

 the bark are typical rather of species than of groups, and that, among shrubs or small stunted or scrubby 

 trees, the cortical character is of very little avail, even for the discrimination of species . . . .' 

 (p. 186). 



The correlation of anthers and barks is interesting. 



This Series, No. V, is subdivided into nine Sub-series, viz. : — 



1. Subsessiles. 6. Subexsertse. 



2. Recurvte. 7. Inclusa?. 



3. Robustee. 8. Corymb osse. 



4. Cornutse. 9. Eudesmise. 



The characters of these Sub-series (with the incidental references to stamens in 

 Nos. 4 and 9, viz., Cornutee, " Stamens erect or flexuose in the bud, but not inflected," 

 and Eudesmise, " Stamens sometimes in four clusters "), are founded on flower 

 attachments, position of the valves in the fruit, and the nature of the fruit itself, and 

 need not be further discussed at this place. 



