626 THE AMERICAN NATURALIST. [VoL. XXXIV. 
P. vulgaris we get a curve of type IV, showing considerable 
asymmetry of its slope, probably in consequence of the ab- 
normal frequency of the zero variant, while the empirical poly- 
gon of variation apparently corresponds to a nearly symmetrical 
curve. Of course there is no agreement between these curves 
and the empirical series of variation. 
The correlation between the numbers of dorsal and ventral 
spines in P. vulgaris is positive, as is the rule in antimerically 
arranged homologous characters; its coefficient r= .3878 € 
0177. 
Besides variation in the number of spines there are some 
more individual differences in the shape of the rostrum of 
P. vulgaris, which will be seen from Figs. 2-27. The total 
rostrum may be compared with a knife, the blade of which 
is serrated on both edges, its handle bearing spines only 
on the back. In the spaces separating the spines there 
are feathered hairs in a single series; from the smooth ven- 
tral surface two rows of hairs extend from each side down- 
ward, like the rafters in a roof. In the males (Figs. 2-5) the 
rostrum is generally more slender and its handle longer tban in 
the females (Figs. 6—27). While in P. varians Weldon found 
the apex of the rostrum bifid in more than half (52.8%) of 
all the individuals investigated, in our P. vwgaris it always 
was single-pointed. 
With increasing numbers the spines stand together more 
closely and extend more forward to the apex of the rostrum 
(cf. Figs. 17 and 19 with Figs. 6 and 7) ; in case of very high 
numbers (extreme variation) they are sometimes irregularly 
arranged and of different sizes (Figs. 14—16, 18). Spines are 
rarely bifid (Fig. 13, third ventral), reduced (Fig. 20, between 
second and third dorsal; Fig. 21, behind first ventral) or 
entirely absent in a part of the otherwise normal rostrum 
(Fig. 22, between second and third dorsal) Very frequently, 
however, especially among the smaller numbers of spines, one 
meets with curious malformations (Figs. 5, 23-27); apparently 
due to regeneration, the rostrum, on account of its exposed 
situation, suffering easily from traumatic injuries. It would 
be interesting to learn if in the course of several moltings the 
