JUNE 14, 1901.] 
respect to the number of anal plates and 
genital pores. The other figures all repre- 
sent different cases, none of them exactly 
duplicated in the whole collection of sixty- 
five. In No. 1, the mode, we recognize 
four anal plates 2n anterior, a posterior, a 
right and a left. Five genital plates, meet- 
ing by a considerable distance and totally 
excluding the ocular plates from the border 
of the anal ring. Each genital plate is per- 
forated by a single genital pore. A study 
of the cases shows that none of these points 
are constant. 
First, as to the number of anal plates, we 
find variation between three and five with 
four as the mode. In the case of those 
with five plates, two have the fifth very 
small, barely distinct, and the position of 
each one of the four large ones is similar to 
that of the mode. It is interesting to find 
that the fifth in both of the two cases is in- 
troduced at the same point, viz., near the 
left anterior radius. The position of the 
plates is nearly constant in the specimens 
with four, though there is a little shifting 
usually toward the right. Thus in No. 
4, the centers of the anterior and poste- 
rior plates are strictly in the axial line, 
butin No. 7they arenot. Inthe cases with 
three plates, the anterior seems to have 
been the one to disappear, the posterior 
being present and showing the same tend- 
ency to shiff as just noted. In the cases 
with five equal plates, Nos. 14-17, the 
plates show a tendency to be all of them 
radial in position, 7. ¢., alternate with the 
genital plates. 
The number of pores in the genital plates 
is also subject to variation, which occurs ina 
greater percentage of cases than the preced- 
ing. In the mode, No. 1, there is a single 
genital pore in each genital plate. In No. 
2, the right posterior genital plate has two 
pores, one being smaller. In No. 3, there 
are two large pores well separated. In No. 
19, there is in the right posterior genital plate 
SCIENCE. 
939 
a pore imperfectly divided into two ; itis the 
least deviation from the mode in the collec- 
tion. No. 5 is an interesting case in which 
in one plate there are two pores nearly 
separate and in another three, one distinct 
and the other two fused. In No. 8, one 
plate has two pores and another three. 
Three pores is the maximum number that I 
have found in one plate. In some eases all 
three pores are united as in No. 9, or one is 
distinct and two are united as in 7, or all 
are distinct asin 7, 8,10. The maximum 
number of genital plates in one individual 
thus furnished with extra pores is four 
(Nos. 10 and 11 and, possibly, 9); in no 
case are all the genital plates thus varied. 
As to frequencies, I notice that it is more 
often that we find two pores than three. 
No one of the five genital plates is wholly 
exempt from this division of the genital 
pore, but the madreporic plate is least fre- 
quently variable. I have found two cases 
in sixty-five in which it had two pores, and 
one case with three. 
Summarizing the facts as to the number 
of genital pores, we get the following table, 
in which the number of pores for each plate 
is shown ; thus the upper and left block of 
figures shows that the left anterior genital 
plate has one pore in seven cases, two pores 
in five cases and three in no case, and so on 
for the remaining four plates: 
Wibb ou Ge =7 111111111—9 
22222 =5 22 =z 
= =0 3 =! 
11111 =5 iil = 
222222 —6 2222222=7 
3 =! 3 I 
11111=5 
222 =3 
3333 —4 
The three posterior plates are thus shown 
to be more variable than the two anterior 
plates, and the posterior plates show a 
stronger tendency to having three pores 
than any of theothers. It is interesting to 
note that in the subcarboniferous sea- 
