1897 - 98 .] Dr Masterman on Ar chimeric Segmentation. 307 
show many archi-coelomate characters. Lastly, from forms pos- 
sibly somewhat allied on the one hand to the ancestors of the 
worms, and partly to the Endojprocta , the Mollusca have arisen. 
These have not become metamerically segmented ; they retain the 
metacoeles as simple coelomic sacs with nephridia, but the pro- 
tocoele and mesocoeles are progressively broken up. 
In 1876 Huxley wrote “ A mollusk appears to me to be essen- 
tially an Annelid which is only dimerous, or trimerous, instead of 
polymerous ” (loc. cit.). In this respect the Mollusea are archi- 
coelomic but their high differentiation causes them to differ from 
almost every other feature of the Archi-coeloinata, so that they 
must be regarded as having progressed above this group, but 
without the aid of metameric segmentation. If the presence of 
archimeric segmentation without the metameric be the sole criterion, 
then they would rank as Ar chi-coelomata. 
The relationships of the Arcld-coelomata may be indicated by 
the accompanying Table, and they would be further expressed as 
follows : — 
Triploblastica. 
1. Pseudocoela. 
2. Coelomata. 
1 . 
2 . 
3. 
4. 
Archicoelomata. 
A. Echinodermata. 
B. Archi-chorda. 
C. Chsetognatha. 
D. Brachiopoda. 
E. Ectoprocta. 
E. Endoprocta? 
G. Botifera? 
H. Sipunculoidea. 
Annulata. 
Mollusca. 
Eu- chorda. 
