THE METAMORPHOSIS OF ECHINODERMS. 79 
determined, and I have therefore confined my figures to this 
stage. 
Their spines are of two different kinds, and are of great 
assistance in identifying the different series of plates, the 
general arrangement of which can be made out from figs. 34 
and 35, and from the ventral view given by Lovén (11, pl. xvii, 
fig. 149). On the aboral side of each primary tentacle lies a plate 
with two quadrangular spines, such as we have seen on the basals. 
In my latest stage (fig. 34) they curve round as if to embrace 
thetentacle. Alternating with these, and usually rather nearer 
the oral surface, are five interradial plates each bearing one 
spine of quite a different pattern, with six longitudinal rods 
instead of four (fig. 33). Théel (35, fig. 99) has given figures 
of the development of these plates in Echinocyamus, which 
agree with what I have observed in Hchinus, though rather 
too diagrammatically regular. Several of the other plates, 
the position of which need not be described, bear spines of the 
same pattern (see figs. 84 and 35)—indeed those over the 
primary tentacles are the only plates developed on the left 
side which have spines of the quadrilateral pattern. 
All these plates and spines have been carefully studied, both 
in whole specimens and by means of maceration and dissection. 
In my oldest larva, however, an accident prevented the use of the 
latter method, and it is therefore possible that I have exaggerated 
in fig. 384 the amount of curvature existing in the plates overlying 
the tentacles; I do not think, however, that this is the case. 
This (fig. 84) is the latest stage to which I have been able to 
rear, from the egg, the young Echinus microtuberculatus. 
The next stage (still of the same species) which I have been 
able to obtain is very much older, and a considerable gap is 
left between the two. Being unable to see the plates clearly 
in the whole animal, I cut off the aboral portion and examined 
it as a transparent object, the result of my observations being 
shown in fig. 86: many of the spines are broken, and the 
plates at the margin have been much injured in the process of 
section, but still the figure shows much that is interesting, 
The basal plates have already (5) been traced from the plates 
