CoraUnin in F.forbesi, f<(j: occide/i talis. — Girly. 135 
tions, while others sometimes accompany them, but without 
system or order. 
In the development of these eorallu, two elements can be 
readily distinguished, viz.: the initial cell with the four buds 
which it produces, making a well-defined group: and the true 
peripheral and interstitial buds, whose number and position 
are largely dependent on the number and position of the first. 
The character of the former group may xtxvy in different spe- 
cies of Fai-ositas. The secondary corallites may be more or 
less than four, and may be differently disposed about the 
initial cell, but the law governing the introduction of the 
other buds is both obvious in its nature and apparentU' the 
same throughout the genus. In Furosites forhesi the four 
secondary buds seem to form a (!onstant increment of aug- 
jnentation which each indivitlual in the colony tends to per- 
petuate. Several instances have been noticed, one of which 
is figured on plate viii, figure 21, where a polyp, separated for 
some reason from the rest of the corallum,has produced four 
buds from its dorsal side in the same manner as the prim- 
itive corallite. All the so-called irregularities have been man- 
ifestations of the same tendency. That this number is in 
most instances reduced to one or two in crowdetl colonics 
seems to be due to the rapidity with which one generation suc- 
ceeds and closes over another. 
In the appearance of the first four buds, no one order pre- 
dominates. The four possible arrangements have all been 
observed, and in nearl)'' equal proportions. A few instances 
have been met where the initial corallite developed but three 
buds in the second series. After a careful inspection of all 
undoubted examples of this occurrence (plate vii, figure 20) it 
is evident that the two outer buds are the oldest, and that all 
three are of unusual size. The space available for growth 
having been thus defined and partially occupied b}^ tiie two 
outer cells, and the inner bud, the first to appear, having 
usurped the rest, the remaining bud was forced either not to 
be developed, i. e., to appear as a pore, or to come in as an in- 
terstitial cell. 
In regard to the parentage of the peripheral liutis as a whole, 
there can be no (luestion, inasmuch as they arc by ilelinition 
external, antl tiieir origin and growth can be made out on any 
