402 MR. F. DABWIN ON THE GLANDULAR RODIES ON 



nate in constricted points composed of cells differing slightly in ap- 

 pearance from those of the rest of the leaflet ; now, although the 

 points on thelower leaflets which are destined to become food-bodies 

 soon outgrow those on the upper ones, yet there can be no doubt 

 that they are morphologically identical with one another. Food- 

 bodies are usually developed on the six or seven lower pairs of 

 leaflets ; but sometimes, by an arrest of development, one of these 

 terminates in a simple point hardly larger than those on the upper 

 leaflets. This is seen in fig, 2, where p ought normally to have 

 grown into a body like/ 3 . In accordance with the principle of 

 compensation of growth, the leaf is frequently dwarfed in cor- 

 respondence with the development of the food-bodies, as for 

 instance at f l in fig. 2 ; occasionally the leaf proper is quite 

 aborted, and the food-body is attached to the petiole simply by a 

 minute stalk. The question next arises, "With what structures in 

 other plants are these points homologous ? If we imagine the bi- 

 pinnate leaf of the Acacia converted into a pinnate leaf by the 

 coalescence of the pinnae, the tips of the old pinnae will become 

 the teeth of the new serrated pinnae. Now, withered tips are 

 found on the points of the serrations of many leaves ; and Beinke 

 has shown that such points are the remains of glands which are 

 highly developed at a very young stage of growth, and die off as 

 the leaf grows*. I conclude therefore that the food-bodies are 

 homologous with the serration-glands of Eeinke. 



Development of the oil. — In the youngest condition in which I 

 have been able to examine the leaflets, i. e. when the food-bodies 

 are only -05 millim. in length, no oil is visible ; there is merely a 

 cell-sap cavity surrounded by yellowish protoplasm. I cannot be 

 certain at what age the oil first appears; in food-bodies a little over 

 a millimetre in length (a full-grown one measuring about 2 millims.) 

 a considerable quantity of oil was found. In these specimens 

 there was a cavity in the cell which contained no oil ; nor were its 

 walls smooth and rounded like those of the cavities in which oil 

 is embedded in a protoplasmic matrix, but showed rather a jagged 

 outline. This cavity of course diminishes in size as oil is deve- 

 loped in larger quantities in the protoplasm enclosing it ; possibly 

 oil globules may escape into it like a secretion into a duct. 



In sections taken from a young food-body about a millimetre in 

 length, the protoplasm enclosing each cell-cavity is seen to contain 



i 



* Gottingen Jfachriekten, 1873, p. 825. 



