Covalbnii in F./or/)es/\ rar. nccidottah'.s. — Girfi/. 139 
lowing explanation of these two elements ( / and .^) and their 
origin is suggested : 
An initial corallite which gives oft' in succession six buds 
equally distributed around its perimeter is taken as an arche- 
typal form. Then, following the mode of growth observed 
in Favositcs forhi'si^ before the secondary corallites are large 
enough to bud, the divergence of these cells allows six inter- 
stitial buds in the six corners, thus decussating with the sec- 
ond generation. 
In the third generation, the inherited tendency would be for 
the six members of the second generation each to produce six 
buds. The diagram (plate viii, figure 21) represents the con- 
dition of the corallum at this period. It will be seen that, 
owing to the interstitial buds, one side of the corallite {b) is 
appressed to the corallum. As a result but four buds can be 
developed normal!}'" by each individual, the two others exist- 
ing in potentin as pores, since pores and buds are fundamen- 
tally homologous. This process is repeated by each member 
of the corallum (except the^ initial cell) and at each act of 
gemmation. By accelerated heredity there would finally re- 
sult an initial cell giving otf f(jur buds on one side of the 
polyp. 
In the suppositious corallum above suggested, the initial 
cell was assumed to put forth six buds, constituting the sec- 
ond generation of the corallum. A little later six others are 
produced, alternating with the first series and forming the 
first set of interstitial cells. The third whorl of buds would 
be above the first, the fourth above the second and so on, 
forming in this way twelve vertical rows of buds or pores. 
This fact shows a striking and significant identit}^ in number 
with the almost invariable number of septa, and therefore of 
interseptal locLili observed in Farosifcs. Moreover, in F./orhrsi 
the initial corallite at the first generation gives oil" regu- 
larly four buds, then three others alternating with tliese and 
so on. This development, however, is on one side only, and 
if continued completel}'^ around the circle W(»uld give twelve 
radii of gemmation (see diagram 22, plate viii). In this set 
of coincidences rests the defense for choosing precisely six 
secondary buds rather than any other number. 
That the regular hexagonal form of tJH' cttraliitf rules 
