18 J. D. Dana on Zoophyies. 
near its base and give rise to branchlets, thus commencing to be- 
h 
lose the budding power, and the branch commences to fork or 
subdivide into two branches. This mode of branching ( furcato- 
ramose) characterizes the Porites, Sideropore, Pocillopore: and 
other speci . 
67. If the polyps of a parent cluster rapidly elongate as in the 
Gorgonie the cluster does not widen, and such species cannot 
branch by furcation. 
68. The position of branches as well as theifsize is determin- 
ed by the principles adduced. In Madrepores the angle which 
the polyps make with the axis of the stem is the angle with 
which the new branch begins, and this angle varies little there- 
fore in the same species. ‘The length or size of the polyps, and 
the breadth of a budding cluster, limits the diameters of branches. 
69. In the horizontally growing Madrepores, the new branch- 
lets form on the outer or lower side of the branches, and after- 
wards become successively nearly or quite vertical. This mode 
of budding retains the zoophyte in a horizontal position. 
The warty prominences ona Pocillopora arise from the fact 
that certain small clusters of polyps of two or three in each, and 
regularly distributed, continue to bud for a while among those 
which from age are just leaving the terminal clusters. 
70. B. Superior or Terminal Budding. In lateral budding, 
the prolate growth of polyps takes place by the extension of their 
inferior portions; while in terminal budding it proceeds from 
extension of the summits. The following figure of an 
Echinopora shows well this peculiarity, and it is still better 
understood on comparing it with the Gemmipora, figure 33 
Shiety 
Echinopora. 
The margin grows by extension of the wpper parts of the poly: 
instead of the lower, and it is connected with terminal bdo 
The cells therefore are not united below to one another as in 
the Gemmipora. The new bud opens in the extending margin a 
