44 DE. J. P. GEMMILL ON THE DEVELOPMENT OF 



poles retain the peculiarity of colouring which marked these poles in the egg itself. 

 As development proceeds, the yolk-granules undergo gradual intracellular digestion. 

 The process continues for a remarkably long time, which it will be convenient to 

 divide into two periods : (a) up till the beginning of metamorphosis and (b) during the 

 middle and later stages of metamorphosis. 



(a) At the end of the free-swimming stage many granules may be demonstrated 

 with the help of osmic acid in all the tissue-cells except the epidermal ones forming 

 the sucker and the adhesive tips of the larval arms. On the whole, they are most 

 numerous in parts which have as yet undergone little growth, or in which much 

 growth has still to take place — e.g., lining of enteron, of hydroccele, of portions of 

 posterior ccelom, and epiderm over posterior part of body. But it is within the 

 mesenchymal cells, particularly those of the disappearing preoral lobe, that the yolk- 

 granules are now to be found in greatest abundance, and here their size and staining 

 capacity is still practically undiminished. During the course of metamorphosis, the 

 granules disappear from the cells of the tissues generally, but in the case of the 

 mesenchyme this does not take place till the very end of metamorphosis after the 

 mouth-opening has formed. The enteron also shows certain striking peculiarities in 

 this respect which are referred to below. 



The mesenchyme thus serves an important function in the development of the 

 highly-yolked Solaster egg, by retaining for the later stages a reserve store of nutrition. 

 The diffuse distribution of this tissue will render its store of nutriment accessible to 

 the cells of the various actively-growing structures. As regards the origin of the 

 mesenchyme, it will be remembered that at the end of gastrulation only a very small 

 amount of this tissue is present between the epiblast and the primitive hypoblast. 

 The amount increases very rapidly during the early differentiation of the three 

 chambers into which the archenteron becomes divided. The increase takes place 

 through budding from the basal aspect of the cells lining these chambers. It occurs 

 earliest and most abundantly from the lining of the anterior coelom, and partly 

 accounts for the flattening of the walls of this cavity. Less abundant and later is the 

 production of mesenchyme from the lining of the posterior ccelom. Little is 

 contributed by the enteron, as might be expected considering the small original size 

 of the chamber and the very great expansion which it afterwards undergoes. The 

 contents of the young mesenchyme-cells consist in chief part of large-sized yolk- 

 granules. In this connection it is interesting to compare with Solaster the very 

 heavily-yolked Starfish eggs from the Franklin Islands described by E. H. Henderson 

 (ii). These specimens were in metamorphosis, and the type of development 

 corresponded exactly to that shown in ABterina as described by Macbride, the chief 

 difference being that the embryos under discussion were much larger, and had all the 

 interstices of the body gorged with yolk to such a degree that it formed nine-tenths 

 of the whole bulk. The yolk was made up chiefly of large globular or irregularly- 



