XCIV 
THE VOYAGE OE H.M.S. CHALLENGER. 
In the next place the researches of Barrels, Kent, and Heider make it clear that the 
characteristic choanocytes appear in the larval Sponge at an early stage of development 
(in the blastula). In no larval forms of the Ccelentera or Turbellaria have similar cells 
been observed. 
Further, while the Sponges are distinguished by the presence of choanocytes, the 
Ccelentera are equally distinguished by the possession of cells bearing nematocysts ; I 
do not know whether any Ccelenterate has yet been described in which nematocysts are 
not present ; it is true that these structures are also present in the Turbellaria, but that 
only points to a common ancestry for this group and the Ccelentera. 
If we derive the Sponges and Ccelentera from a common Metazoic ancestor, the 
simplest hypothesis will lead us to regard this as possessing both choanocytes and 
nematocysts ; as the two groups diverged, one might be supposed to lose the choanocytal 
character and the other the nematocysts. I do not deny that this is possible, but it 
seems to me that a simple and wider reaching explanation may be found. 
It is generally admitted that it is to the Infusoria that we must look for the origin 
of the Metazoa ; if now we turn to this we find already existing collared fiagellate 
individuals having a remarkable similitude to the choanocytes of the Sponge, and ciliated 
individuals possessing nematocysts and resembling the ciliated nematocyst-bearing cells 
of the Ccelentera and Turbellaria. The genera in which trichocysts or nematocysts 
have been observed are Epistylis and Bursaria among the Ciliata, Polycricos among the 
Dinoflagellata, and in the Microsporidian division of the Sporozoa. The Choanoflagellata 
do not possess them. The genera just mentioned as possessing nematocysts are so 
related that there is no reason for supposing that the nematocysts have independently 
originated. 
The conclusion to which so far I am led is, therefore, that the Sponges have 
their origin in the Choanoflagellata, and the Ccelentera in Ciliata furnished with 
nematocysts. 
It may, of course, be suggested that if the Choanoflagellata have arisen from the 
Ciliata, then both Sponges and Ccelentera might have been produced somewhere about the 
point of transition, and it is quite conceivable that this might have been a point at which 
some large ciliate cell remained attached or in continuity with choanoflagellate cells to 
which it gave rise, and in this way we might attempt to explain the early distinction 
of the choanoflagellate cells from the large granular cells in the amphiblastula of the 
Calcarea (Megamastictora). It will be found on consideration, however, that this view 
presents serious difficulties. 
For the purposes of this Eeport the Sponges will be regarded as the only phylum of 
the separate subkingdom Parazoa — 
