7 



part and, therefore, the above-rindicated division, I believe, has no 

 fiandaiaental importance. On the contrary, it remains ae a passive division 

 with the bipartition of tb^ plasma after the manner followed in this case. 

 In fact, during nucleetr division the elaioplast remains inert and almost to 

 the moniBnt in which Uihe new membrane which begins to form comes against it. 

 Then it bends and takes on a semilionar form. The extremity of the thin membrane 

 forms and then splits into incurved points generally in two unequal parts. 



With increase in size of the cell the elaiopla^jb^ grows in volume and 

 at complete development exceeds the dimejaftiions of the nucleus, becoming often 

 four or five times larger than the latter. Later the elaioplast of the epi- 

 dermal cell, where it is found at a certain distance from the extremity of 

 the apex of the floral axis, begins to undergo phenomena of degeneration. 

 In fact, it diminishes in voliane, becomes impoverished in oily substance, and 

 presents some scattered vacuoles in its jnass among which often there is to 

 be distinguished a large central vacuole. Finally in the cell, as development 

 advamces, nothing more than a residutan of the elaioplasts can be found, and 

 freijuently also this comes to be completely reabsorbed in the cytoplasm in 

 which it is immersed. 



It is to be noted that in this species the elaioplast fismains almost 

 always throughout its life close to the cell nucleus. 



In the species in question the elaioplasts are encountered in the 

 epidermis and in the parenchyma of the perianth, of the floral axis and of 

 the peduncle, of the stamens, of the pistils, and of the young leaves; they 

 are absent in the ovule, in the seed, in the adult fruit, and in the roots. 



