930 HIROSHI OHSHIMA 
of the first order (P, 43) arranged in a dextrorse spiral, and with 
an angular divergence of one-quarter or two-sevenths. Of 
these, no. 2 (P,) is the largest, being beset with eight to ten 
branches of the second order (S;.,). The latter are arranged 
in a sinistrorse spiral except on no. 1 of the first order (P,), 
where the arrangement is dextrorse. In some comparatively 
larger ones of the second order, one can distinguish two to three 
branches of the third order (7, 5). 
In the two ventral tentacles the features are quite different 
(Text-fig. 10,8). These keep for a considerable period a very 
simple appearance, in that the tip is branched twice dichoto- 
mously. This may probably be an adaptive change. The left 
branch undoubtedly gives rise to no. 2 of the first order, which 
grows as large as the main stem rising from the right branch. 
They later give out branches along their whole length as seen 
in the adult state. No. 1 of the first order appears later on the 
outer side immediately below the bifurcated pot. In none 
of the other Cucumarids does such a peculiar feature seem to 
have been noticed. . 
Mitsukuri (ante, p. 175) first noticed the regularity of 
branching of the tentacles in that the ‘ pinnules’ stand in a 
spiral arrangement (sinistrorse as judged from his figure), with an 
angular divergence of one-quarter, and that the second pinnule 
is the largest. But as regards the direction of the spiral his 
statement does not agree with my observations. Ludwig 
(22, p. 185; 24, p. 97) stated that both in C. planci and 
Phyllophorus urna the five primary tentacles first bifur- 
cate at the tip, and then each branch produces side branches. 
In C. echinata I observed no such terminal bifurcation except 
in the ventral pair (Text-fig. 7). Kowalewsky (1%, p. 6) 
was of the opmion that the -branching of the tentacles in 
CG. kirchsbergii occurs, not simply from terminal bifurea- 
tion, but from producing a bud near the apex of the tentacle. 
Inthe ten-tentacled stage of PSPeudocucumis africanus 
of about 6-5 mm. in length, no such differentiation of the ventral 
pair is found, all being beset with several side branches. 
Increase of Pedicels.—The order of the appearance 
