136 Tlie Aniericd)! Geohxjist. Marcli, 1895 
wcll-prescrvod spccinu'ii. It niiglitbe asked, liowevcr, wliethor 
the buds which ultimately surround the initial cell ai-e not 
immediately given off by it. This does not appear to be the 
case, both from the long int(;rval which separates their ap- 
pearance from that of the first four, as well as from the gen- 
eral law which governs the introduction of new cells. More- 
ovei", a partially developed corallum was tlissected by the 
writer, by means of acitis, and the basal pores of such buds as 
were developed were found to pass into cells of the series h 
and c, and not into the initial cell. 
The number of interstitial cells has already been seen to 
var}' considerably, depending as they do upon the position 
and development of the other corallites. Their own position 
seems to be due to the general crowding of the colony. Those 
cells which are interstitial between the first and second gen- 
eration, spring from the first, those between the second and 
the third from the second, and so on. Not only has this been 
observed in the single specimen dissected, but it also agrees 
with the strong unilateral tendency which is exhibited in 
other ways. In general, it may be asserted that the oldest 
cell buds first, and that the cells of other generations send off 
buds in the order of their 5>ges. Thus the order of the whole 
corallum is in a measure predetermined by the order of the 
four buds constituting the second generation. The period at 
wliich the initial cell is completeh^ surrounded varies slightly 
in different coralla. It is evidently dependent on the rapidity 
with which the enveloping buds, especially the final pair, in- 
crease in size. 
Specimens showing stages in the growth of Fdrosifcs 
forhcsi are figured on plate vii, figures 1-2(3; plate viii, figures 
28, 24 and 25. 
Farosifex s/tiiiiffcriis Hall. 
This species is less abundant, and less satisfactory for study 
than h^trosHes forhcsi, although obtained from the same lo- 
cality and in the same preservation. In its development, it is 
closely allied to the species above discussed. Ilie initial cell 
produces four buds from the dorsal side, as in F. forbcsi, but, 
through prolific budding, these are generally separated by a 
large number of interstitial cells. Subsequent growth is more 
dorsal than in /''. /orhr.si^ so that, in all the specimens ex- 
