Flora oj AuMralia. 137 



Variety GuNNii (P. Gunmi, r>entli.). In Ann. der Wien. 

 Mus., ii., 82 (1839). 



This has naiTOwer, usually spreading stipules, the inner larpje 

 bracts are usually absent, but in all the forms, including the 

 type, the bracts usually fall as flowering advances, and some 

 forms of our Gunnii, have much larger' bracts than others. In 

 the typical forms the leaves are usually broader at the base and 

 taper more or less towards a usually pointed apex. In the 

 typical A. stricta the leaves are usual^^ broader near the apex' 

 then suddenly contracting to a distinct point. The difference 

 is more constant on the larger t^tem leaves. 



Variety Larc.iflohens (P. laugiflorens), F. v. M. 

 In Fzienth. Fl. Austr., ii., 134 (1864). 



The flowers may be either axillary or lateral, or in terminal 

 clusters, the bracteoles are usually inserted higher up on the 

 calyx tube, and the two upper calyx teeth are less or not at all 

 falcate, and the fruits usually smaller. 



Variety Ixcurvata, new var. Locality, Frankston, Coll. 

 J. W. Audas, 1907. 



This has the leaves with hard, minutely-pointed, recurved tips, 

 giving the plant a peculiar harsh feel when drawn through the 

 fingers. In some respects it is intermediate between the variety 

 Maideni and the type form. 



Pultenaea retusa, Sm., comes near to some forms of P. stricta, 

 l)ut the calyx teeth are of more equal shape and length, and the 

 calyx is hardly bilabiate. The usually straight upper calyx teeth 

 of variety largiflorens show an approach to this species. 



It may seem a bold course to reduce these three well-kno-\\Ti 

 species, but the numerous connecting links leave no other course 

 possible, and there is no evidence as to the existence of hybridi- 

 zation between these four species. Variety largifloi'ens, shows 

 the largest, variety Gunnii a lesser, and variety Maideni the 

 least divergence from the type, but the same reasons that could 

 be urged for their maAntenance as distinct species could be used 

 to found at least 12 species out of the numerous connecting forms. 

 It may be taken as a general rule that in all large genera the 



4 



