176 THE INSTABILITY OP LEAF-MORPHOLOGY 



for the rigidity of the description. In following a botanical de- 

 scription with herbarium-material, frequently limited in quantity, 

 and from few, and often obscure localities, the variations in 

 foliage brought about by local conditions must, in the absence 

 of such information, be lai'gely speculative. Nor does the pos- 

 session of a type-specimen entirely remove the difficulty, unless 

 the environmental conditions under whieh the type existed are 

 available. The most valuable assistant to the taxonomic botanist 

 is personal observation of his flora in situ, but as this is fre- 

 quently impracticable, the carefully compiled field-notes of com- 

 petent observers, with a knowledge of the factors liable to affect 

 the morphology of leaves, are extremely helpful. In this direc- 

 tion, the value of the numerous references embodied in the 

 writings of local botanists (who are, without exception, field- 

 botanists) to the influences at work on the alteration, in our 

 native vegetation, of the characters relied upon by the systemat- 

 ist, cannot be overstated. The following expressions of opinion 

 may be quoted. Bentham (4; iii., p. 186), in his remarks on the 

 genus Eucalyptus, says : " but to the botanist who is unable to 

 compare them in a living state, the due limitation and classifica- 

 tion of their species presents almost insupei'able obstacles." In 

 his Presidential Address to this Society (8; 1915, p. 649), Mr. 

 Henry Deane refers to the determination of plants from leaves, 

 and quotes a Presidential Address to the Linnean Societ)^ in 

 1870, by Bentham, who points out the unreliability of determina- 

 tions made on leaves only, and mentions that De Candolle had 

 been in error as to Natural Orders of species of which he pos- 

 sessed leaves alone. Later (9; 1900, p. 581) Mr. Deane stigma- 

 tises as ridiculous the case of a botanist who would attempt a 

 classification and description of a hitherto unexplored flora on a 

 collection of leaves, and notes (p. 588), that Ettingshausen points 

 out in his paper the impossibility of carrying out any system of 

 classification on leaf-characters, offering examples of similarity 

 in widely separated, and heterogeny in more closely allied plants. 

 Mr. Deane in his paper(9), also tabulated a number of species, 

 and supplied figures of leaves, instituting comparisons as to their 

 similarity in opposing, and variation in allied plants, with special 



