1879.) } 299 [Fewkes. 
respect from the tentacle itself. They arise as simple buds from the 
tentacles, and in early conditions are sessile. It is only when more 
developed that the secondary stems are formed. In the first stages 
we have a club-like body, slightly constricted at its base and position 
of union with the tentacle. In this primitive knob there are many 
large lasso cells, always imbedded in the surface layer. These cells 
are described by Gegenbaur. 
The first differentiation in the development, which up to this time 
‘presents no variation from that of other Siphonophores, is the aggre- 
gation of these cells at one pole, which is the distal extremity of the 
knob, and the formation of a finger-like extension, tipped by a cluster 
of smaller cells; this cluster of smaller cells is never lost, and forms 
one of the characteristics of the first kind of tentacular knobs. They 
may be seen in fig. 4, where the prolongation is greater than in sub- 
sequent stages. 
From both sides of the under hemisphere of the knob a bud now 
forms, which is simply a proliferation of the walls of the appendage; 
fio. 5 gives an idea of the appearance of the knob at this stage. 
Lasso cells are scattered irregularly through these parts, which are 
smaller than the first formed, although larger than the cluster men- 
tioned as existing at the apex. This is very near the stage which 
Gegenbaur regarded as the first form of tentacular ena; its devel- 
opment, however, is not yet finished. 
The two side appendages lengthen, become more slender; the me- 
dian portion becomes more spherical, and a well-defined neck forms 
at its base. ‘The lasso cells, which were formerly irregularly scat- 
tered over the whole surface, aggregate into two clusters, one ter- 
minal and the other about midway down the arm. Various drawings 
were made of this stage, of which figs. 5, 6, 8 give a good idea. 
This is the most developed condition in which I have found the first 
kind of tentacular knob. 
There is a remote resemblance between this knob and that of the 
genus Agalma. The envelope so well known in this genus is, how- 
ever, wanting. It would be very hard, in the present stage of 
our knowledge, to carry out such an homology, which may be only 
superficial, but it would bea convenient way to designate this kind 
of an appendage, from others yet to be mentioned, if it were known 
as the Agalma-like knob. 
Very different in the method of development from that of those 
already mentioned is the history of the second kind of tentacular 
