BY R. T. BAKER, 379 



preferred to refer the species under consideration to the Section 

 E u p r o s t a n t h e r a. I am influenced in such a decision by its 

 mode of inflorescence as well as hy the fact that one or two 

 species with only rudimentary appendages are already included in 

 this Section. 



In the species of Section Klanderia of Prostanthera the 

 corolla tube is so very distinctive, being " narrow at the base, 

 usually incurved and dilated upwards, the upper lip erect, concave 

 or arched, the lower lip shorter or at any rate not longer and spread- 

 ing," whilst in this species the corolla tube has the lower lip 

 longer than the lobes, is not incurved or narrow at the base, — 

 points that would not justify its being classified with this Section. 



Neither can it be included under any of the species enumerated 

 under Bentham's Series C o n v e x se and S u b c o n c a v se, as all 

 those have axillary flowers and anthers with one appendage about 

 twice as long as the cell. 



Of the species described under Eu prostanthera it most 

 resembles P. rotunclifolia and P. violacea in its close terminal 

 racemes, but differs from them in the form and size of its leaves, 

 shape of corolla, and, of course, virtual want of anther appen- 

 dages. 



It also differs from P. incana, P. hirtula, and P. denticidata in 

 its leaves being perfectly flat, also in inflorescence, indumentum, 

 and absence of anther appendages; and for the same reason it is 

 excluded from P. rugosa, P. marifolia, P. rhombea, P. spinosa, 

 P. cuneata, P. linearis, P. phylicifolia, P. decussata, and P. 

 empetrifolia. 



Its greatest affinity is perhaps with P. incisa and P. Sieberi, 

 but its leaves are so distinctly or uniformly entire that I prefer to 

 regard it as a connecting link between those two species and P, 

 rotundi folia. From the description of P. incisa one might be led to 

 think it was that species, but when specimens of each are 

 placed side by side the differences are very marked. 



From the above considerations I conclude that in botanical 

 sequence it should come after either P. incisa or P. Sieberi, and 

 be followed by P. rottmdifolia. 



