C. L. Edwards 
263 
small (Fig. 36). With age the holes may become so much filled in with calcareous 
matter as to assume the form of pits (Fig. 41, where 11 of the 15 holes have 
become pits; indicated by the cross lines). Upon some of the plates are ridges 
and then the holes lie in the furrows between the ridges. The standard deviations 
given in Table XX. show that the developed plates are more variable than the 
undeveloped. 
A specimen from Key West, Florida, compared with specimens from Porto Rico 
and St Thomas, presents plates with more holes and at the same time has more 
plates with developmental branches which indicate that yet additional holes will 
be formed. An extended study of the perforated plates from the various localities 
would, without doubt, reveal interesting place modes. 
Theel, 1886, p. 21.5 (cf. also Clark, 1901, p. 258) in describing H. mexicana 
notes that " of the crowded plates two types may be observed, one nwve rounded, 
pierced with minute and commonly more numerous holes ; and the other irregularly 
rectangular, with fewer and larger holes." The first of the two types of these 
authors I consider identical with the developed plate, and the second with the 
developmental stage Type h described in this paper. Liidwig, 1874, in his original 
description of mexicana only mentions "numerous symmetrical perforated 
plates" but figures the two kinds (Taf. VII. Fig. 47a). These facts, together 
with the other points in agreement, establish my claim that H. mexicana Ludwig 
and H. africana Theel are identical with H. floridana Pourtales, and therefore 
synonymous with the last-named species. 
e. Number of Rosettes and Perforated Plates. 
In a piece of the body-wall of 1 sq. cm. area, the average thickness of the layer 
of spicules, extending through about \ of the body-wall, was "82 mm. The 
spicules averaged "04 mm. long x '036 mm. wide x "005 mm. thick, giving a possible 
145,000 spicules in 1 cu.mm., or 11,890,000 to 82 cu.mm., or 1 sq.cm. of surface. 
The spicules are packed densely in this layer, but a discount must be made for 
gaps as well as protoplasmic structures between the spicules. Even allowing this 
to be 50% there would be still over 5,000,000 spicules to a sq.cm. of surface in 
the crowded areas. 
f. Correlation of Rosettes and Perforated Plates with 
Advancing Age. 
In order to determine whether the different classes of spicules are correlated 
with special ages of the Holothurid, 62 from the whole series of 118 specimens 
were carefully examined without, however, any attempt at a quantitative determi- 
nation. Hence the percentages given below are only approximate. A specimen 
usually shows a large majority of either rosettes or perforated plates but, at the 
same time, in each individual some examples, at least, of all classes of spicules 
occur, illustrating more or less completely tlie developmental series from the simple 
rod and x-formed rosette to the complete perforated plate. 
