210 M. B. WELCH. 



can be readily understood since the cavity is more or less 

 ovoid and only a section in the median plane will give a 

 complete view of the structure. 



In the case of the young leaves of E. corymbosa, the oil 

 glands are apparently formed from hypodermal meriste- 

 matic cells and there appears to be no connection with the 

 outer epidermis. It is in this species that a thin coating 

 of rubber is found enveloping the young leaves, and this 

 may possibly be the explanation, as it is apparent that at 

 this stage of development at any rate, no purpose would 

 be served by the gland discharging at the surface. 



The deep-seated glands are characteristic of petioles, 

 e.g., E. Luehmanniana, E. piperita etc., and also of stems 

 as shown in E. Luehmanniana. It is evident that subse- 

 quent cell division in the cortical tissues has occurred, 

 thus removing the glands further from the surface in the 

 older petioles and stems. In the very young structures 

 mentioned the glands approached the surface. It is inter- 

 esting to note that in the petioles of E. piperita the deep- 

 seated oil-glands still contained oil, although no means of 

 direct communication with the surface was found. 



A cross section of the leaf of any of the well known oil 

 yielding species shows a number of these oil glands, or 

 secretory cavities, in which the essential oil peculiar to 

 the species is found. In section the gland varies in shape 

 from an ellipse with the longer axis either at right angles 

 or even parallel to the leaf surface, to an ovate or circular 

 shape, but perhaps the most common form is that of ovate 

 with the apex directed towards the upper or lower 

 epidermis. 



According to Briosi (Anat. Foglie E. globulus), working 

 on E. globulus alone, the first stages in the development of 

 the gland are caused by segmentation of an epidermal and 

 a hypodermal cell. The early stages in the formation of 



