ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 
191 
and not at its outer surface, for neither carmine nor milk-globules 
remain attached to the outer surface of healthy sponges, and the stream 
of water has clearly the function of introducing nutriment and oxygen 
into the interior of the sponge. 
(2) It is clear that the collared cells normally take up the material 
contained in the water that streams through. 
(3) No observation supports the theory of Metschnikofif and Sollas 
that the collared or epithelial cells, filled with food, sink down into the 
intermediate layer. 
(4) Carmine is only rarely found in wandering cells, and it may be 
supposed that such granules as are so found passed in at points of injury 
and not in a normal way. Dr. Lendenfeld does not believe that the 
collared cells give up carmine-grains to the wandering cells. 
(5) With milk, however, it is otherwise ; the globules are taken up 
by the collared cells and then passed on to the wandering cells. 
The method of nutrition of Sponges may, therefore, be thus described. 
The moving flagella on the pavement and (?) collared cells produce a 
stream of water which traverses the canal system of the Sponge, so long 
as it is in a healthy state. Various substances are dissolved and 
suspended in this water. The larger suspended solid bodies do not 
enter the interior of the sponge, as they cannot pass through the small 
pores of the skin ; some, however, do enter by injury of the skin, and 
such are sand-grains, foreign siliceous needles and the like which are 
used by many horny sponges in building up their skeleton. Smaller 
suspended particles such as arise from the decomposition of organic 
substances in water, as well as all substances dissolved in the water, 
enter the sponge, and are all, so far as is physically possible, absorbed 
by the flagellate cells in the chambers. 
The collared cells appear at first to have no power of selection ; this 
is effected by the skin and its pores, which keep out injurious matters ; 
the substances taken up by the collared cells are partially digested, and 
pass, in a more or less assimilated state, into the cells of the intermediate 
layer, which acts as the means of transport for the nutrient material. 
The collared cells excrete what is useless in the food, while the carbonic 
acid formed in the tissue is probably given up by diffusion to the 
surrounding water. 
^he Sponge may be regarded as a living filter which removes from 
the percurrent water, by means of its collared cells, all matters useful 
to them. These cells, in siliceous sponges, have the property of retaining 
the siliceous salts contained in the water; and calcareous salts are 
similarly treated by calcareous sponges. The collared cells of the Horny 
Sponges cannot hold back either lime or silex. 
Although the author’s physiological experiments have not proved the 
existence of a nervous system, they have made its absence more than 
doubtful, for the extraordinary sensitiveness of the skin speaks to the 
presence therein of differentiated sensory cells. The sensory and gan- 
glionic cells are spindle-shaped or pyriform, give off one longer process to 
the surface, or a group of three or more. Aristotle was correct in saying 
that Sponges could contract ; this contraction is the result of harmful 
influences, and is especially observed when poison is dissolved in the 
water in which the Sponge lives — we have here a reflex movement to 
