144 
G. W. Lee — Trepostomata. 
that it is not easy to conceive of it having taken an active part in the growth of the 
zocecium. And one would moreover think that the presence of tabulae would be an 
obstacle to the distribution of the soft living parts. That is, the laminated structure of 
the outer region points to successive periodic depositions of sclerenchyma, and therefore 
to distal growth. 
" But if distal growth be admitted, then it becomes difficult to reconcile it with the 
fact that the respective dimensions of the axial and peripheral regions are constant what¬ 
ever the size of the branch, unless we imagine the intervention of a factor which would 
produce proximally the gradual elimination of the greater part of the wall substance as 
growth proceeds distally ; that is, that there is resorbtion. On the basis of this suggestion 
we may conceive of the growth of a ramose form taking place as follows:— 
T he initial branch begins as a protuberance the diameter of which is within the limits 
of size possible for the species. The accumulation of sediment necessitates an upward 
direction of growth. In order to achieve this end and at the same time accommodate 
themselves to the space at their disposal, the zooecia situated in the centre of the extremity 
are practically straight throughout their course, while the distal end of those that are 
farther from the centre has to be bent at a considerable angle in order to reach the surface. 
That is, the end of the branch tapers owing to the ends of the zooecia diverging from an 
imaginary axis. This divergence creates an increase of space which, as pointed out by 
Ulrich, leads to the distal thickening of the wall and the formation of interzooecial cells. 1 
As growth proceeds, new tubes are interpolated in the central portion of the 
extremity, while the tubes which occupied that position in the stage just considered, have 
to leave their previously straight course and bend in order to reach the surface, at the 
same time resorbing the thickening of their walls in that central position (see diagram, 
plate xiv., fig. 3). From the above assumption it would seem that the branch possesses 
its full diameter from the beginning—an hypothesis substantiated by the fact that when a 
branch divides the secondary branches can be as thick as the mother branch—and 
consequently that size is not an indication of age. 
Besides furnishing a possible explanation of the growth of the colony, the resorption 
hypothesis may throw some light on the raison d'etre of structures otherwise not easily 
accounted for. By assuming that any portion of the axial region was at some time 
situated distally, we can account for the fact that acanthopores are often present in the 
axial region, although their function must have been connected with the surface of the 
zoarium, as pointed out by Cumings. 2 
44, p. 271 ; 45, p. 45. 
2 2, p. 207. 
