Excretory Processes in Porifera. 489 



mesoderm either intracellularly by the loaded clioano- 

 cytes or in some cases perhaps by a j)hagocytic 

 digestion of ehoaiioeyte and its contents. 



4. The immigrated choanocytes aie replaced in the wall of 

 the chambers by cells from the inner tissues, which 

 assume the shape of choanocytes. 



The continual morphological transformation above indicated 

 appears to be the normal process of digestion in, at any rate, 

 tlu' simpler sponges. 



To this statement two objections present themselves : — 



Firstly, Sollas (15) has already, when criticising experi- 

 ments conducted under similar conditions, taken exception 

 to them, remarking that " there is at present no proof that 

 carmine is a food, or that if it is sponges will readily feed 

 upon it." This objection of course holds good with the 

 above in common with all other feeding experiments, and 

 they will be of no value if carmine be proved not to be a 

 tood. 



Assuming it to be not so, it seems difficult to understand 

 why the choanocytes should so very readily absorb large 

 quantities of it. Minute ])articles of sand or other mineral 

 matter if taken at all by the choanocytes must be very rapidly 

 ejected, as their presence in these cells cannot be demon- 

 strated ; at least 1 have, after repeated attempts, been unable 

 to do so. This being so, it is not too great an assumption to 

 su))j)ose that the choanocytes can and do exert a selective 

 power, and are not obliged to absorb and ingest anything 

 and everything which is presented to them, provided it fulfil 

 only the necessary physical requirements ol" size &c. Again, 

 it can be easily shown that numbers of small animals (young 

 starfishes &c., Ruffer(12)) can and do feed upon carmine when 

 it is presented to them, or, at any rate, they behave to 

 carmine in exactly the same way as they do to any nutrient 

 material. 



Lastly, there is not wanting evidence that the carmine 

 particles undergo a certain amount of change in the tissue of 

 the sponge. Lendenfeld (9) finds that the larger particles, 

 after having been expelled from the tissues, have lost their 

 angular outline, and adduces this observation to show that 

 they have at least undergone some change in their passage 

 through the tissue of the sponge. I have also observed that 

 in many of the loaded amoeboid cells there are vacuoles deve- 

 loped around the enclosed particles (fig. 6), perhaps indicating 

 that a process of digestion is going on. 



Taking these facts into consideration, it does not seem to 

 be too great an assumption to suppose that most animal 



