224 
NOTES AND MEMORANDA. 
derm-cells, carrying with them assimilated matter necessary for the 
nutrition of the syncytium, which forms a thick wall beyond. His 
conception of their importance was confirmed by the discovery that 
many of them contain starch. Keller has made an extensive search 
for starch in the cell-elements of sponges, and has found it, or rather 
we should say has obtained the blue reaction with iodine, in cells from 
the following sponges : — (1) Spongilla lacustris, (2) Beniera litoralis, 
nov. spec., (3) Myxilla fasciculata, (4) Geodia gigas, (5) Tethya lyncurium, 
(6) Suberites massa, (7) Suherites Jlavus. The substance, whatever it 
may be, which gives the blue reaction, is not in a granular condition, 
but fluid, and in those cells in which it occurs occupies a large vacuole 
comparable to a fat-vacuole. Neither ordinary nor absolute alcohol, 
nor cold water, dissolve the contents of this vacuole. Keller could not 
find this starch-like substance in Halisarca nor Chondrosia, nor in any 
Calcispongias. It seems desirable in this connection to refer to the 
strictly granular condition in which chlorophyll appears in the case 
of Spongilla, the granules having the form of concavo-convex disks. 
In colourless (etiolated) specimens of Spongilla, the same granules 
are present of a little different form, and as in Neottia and other 
similar plants, these granules turn green (develop into chlorophyll ?) 
on the addition of strong sulphuric acid (see ‘ Quarterly Journal of 
Microscopical Science,’ 1874, vol. xiv. p. 400, where I have recorded 
these facts, and also that of the occurrence of starch in Spongilla, 
though I have not yet been able to find the authority for the latter 
observation, which was made many years previously to Keller’s inves- 
tigation). With regard to the question of the formation of a gastrula 
in sponges, and as to the development of the endoderm of that gastrula 
into the endoderm of the adult sponge, and therefore the continuity of 
the archenteric cavity of the gastrula with the digestive cavity and 
canals of the sponge, Keller has some remarks to offer which do not, 
in point of fact, amount to very much. Like Franz Eilhard Schulze, 
Keller fell into a complete error in his earlier publication on the 
development of calcareous sponges. Haeckel, in his monograph, 
stated that the sponge embryo was at first a hollow one-cell-layered 
sac, on the inner wall of which a second cell layer formed, by delami- 
nation, whilst subsequently a mouth broke through. This was vehe- 
mently denied and ridiculed by Metchnikoff; it was also denied by 
Oscar Schmidt, and by F. E. Schulze, who published a beautiful set of 
drawings showing that after the embryo sponge had acquired some 
thirty or forty cells, one hemisphere of cells became granular and 
enlarged, and then invaginated — sunk into the other hemisphere — 
thus forming a gastrula with endoderm and archenteron by invagina- 
tion. This account was at first accepted as the true one, but it was 
strongly insisted upon by Keller in his former memoir, that the orifice 
of invagination closes up, as in fact the blastopore so usually does 
throughout the animal kingdom, and that the young sponge is then a 
iiiouthless closed sac with two layers of cells. It was in this condition 
that Haeckel saw it and described the further stage in which the true 
mouth breaks through. There is, however, still a great difficulty 
about the development of the gastrula of sponges ; for no one can 
