SECRETING STRUCTURES. 427 



I have also had an opportunity of verifying, and to an 

 extent which I did not at the time fully anticipate, the re- 

 markable vital properties of the third order of secretions, 

 referred to in the memoir to which I have just alluded. The 

 distinctive character of secretions of the third order is, that 

 when thrown into the cavity of the gland, they consist of 

 entire cells, instead of being the result of the partial or entire 

 dissolution of the secreting cells. It is the most remarkable 

 peculiarity of this order of secretions that, after the secreting 

 cells have been separated from the gland, and cast into the 

 duct or cavity, and therefore no longer a component part of 

 the organism, they retain so much individuality of life, as to 

 proceed in their development to a greater or less extent in 

 their course along the canal or duct, before they arrive at 

 their full extent of elimination. 



The most remarkable instance of this peculiarity of secre- 

 tions of this order, is that discovered by my brother, and 

 recorded by him in a succeeding chapter.* He has observed 

 that the seminal secretion of the decapodous crustaceans 

 undergoes successive developments in its progress down the 

 duct of the testis, but that it only becomes developed into 

 spermatozoa after coitus, and in the spermatheca of the female. 

 He has also ascertained that, apparently for the nourishment 

 of the component cells of a secretion of this kind, a quantity 

 of albuminous matter floats among them, by absorbing which 

 they derive materials for development after separation from 

 the walls of the gland. 



This albuminous matter he compares to the substance 

 which, according to Dr. Martin Barry's researches, results 

 from the solution of certain cells of a brood, and affords 

 nourishment to their survivors. It is one of other instances 

 in which cells do not derive their nourishment from the blood, 

 but from parts in their neighbourhood which have undergone 



* See page 429. 



