198 



constant. I have taken advantage of this to form a lateral 

 index, a measurement which I have found useful in differ- 

 ential diagnosis. 



In order to estimate it the total length of the petal is 

 used to form the numerator of a vulgar fraction, and the 

 length of the lateral sepal the denominator. This fraction 

 is then reduced to two places of decimals, and the result is 

 the lateral index. For example, suppose the length of the 

 petals to be 4| lines and that of the lateral sepals 4|, the 



resulting fraction is —^ and the lateral index is 92. 



In order to diminish error it is advisable (when speci- 

 mens permit) to estimate the index in three plants, and then 

 to take the average. 



The lateral indices of species growing in this State are 

 approximately as follow : — 



P. rotundiforuiii, Rogers ... ... 96 



P. eJatvm, R. Br 95 



P. odoratuni, Rogers ... ... ... 95 



P. album, Rogers ... ... ... 92 



P. Avstrale, R. Br 91 



P. pniiiiosm/i , Rogers ... ... ... 87 



P. f/ranle, Rogers ... ... ... 76 



P. occidentale , Rogers ... ... ... 76 



P. roust nctuui, Rogers ... ... 76 



P. patcNS, R. Br 75 



P. Fitzgeraldi, Rogers-Maiden ... 69 



P. f'uscvrn, R. Br 64 



I\ 7)if/ncans, R. Br. ... ... ... 64 



P. Tc'pperi, Mueller-Rogers ... ... 56 



Bidentation of the tips of the lateral sepals, or some 

 trace of it, is almost invariably to be found in all but a few 

 of our species, although, curiously enough, with regard to 

 this Bentham makes the following comment : — "The lateral 

 sepals in two or three instances have been described as 2- 

 dentate. I have never found them so, and believe the error 

 to have arisen either from a slip of the pen, referring to 

 lateral sepals instead of lateral appendages of the column, 

 or the writer to have meant the lip composed of the two com- 

 bined lateral sepals." ("Flora Aust.," vol. vi., p. 335.) I 

 have before me as I write two specimens of P. patens, both 

 examined by Bentham, and in one at least of these there is 

 distinct evidence of bidentation. The incurving of the ros- 

 tellum towards the labellum is a feature which is sometimes 

 mentioned in descriptions as having a specific significance. 

 In the unfertilized flower the rostellum is invariably erect. 

 The incurving is the first step towards fertilization. Thereby 



