508 
HUBERT LYMAN CLARK 
its value. In the first place, Ostergren apparently considers sim- 
plicity of structure as implying a primitive condition ; he scarcely 
refers (except as regards the absence of feet in Apoda) to the sec- 
ondary simplicity often produced by changed habits, and which is 
sometimes called ^'degeneration;" he certainly has failed to give 
due weight to the existence of this factor. In the second place, 
he assumes that the ancestral holothurians were mud-loving and 
mud-eating forms; if these ancestors were worm-like in habit and 
structure this View is tenable, but if they were allied to the regular 
echini, as many zoologists consider probable, it is hard to main- 
tain; certainly in the Echinoidea, it is only among the highly spe- 
cialized forms, the spatangoids, that we find mud-loving and mud- 
eating species. Moreover, there is little doubt that the early 
Metazoa were all plankton-feeders and many Dendrochirota re- 
tain that habit still. It is hard to believe that the class Holo- 
thurioidea had not arisen before the competition for food on the 
floor of the sea led to the use of organic mud as food. In the 
third place, to find the most primitive and ancestral form of 
holothurians in the exclusively deep-water group of elasipods is 
to run counter to one of the principles, which modern oceano- 
graphic work has established ; namely, that the inhabitants of the 
abyssal regions are more or less highly specialized forms, the 
simpler and less modified forms occurring in water of little or 
• moderate depth. 
If Ostergren 's paper errs on the side of overlooking secondary 
simplicity and of definitely asserting his deductions, no such criti- 
cisms can lie against Becher's ('07) exhaustive study of Rhab- 
domolgus. It is difficult to find a theoretical conclusion definitely 
asserted in Becher's work, and it is almost as hard to find what 
characters of Rhabdomolgus, if any, are to be considered primi- 
tive. The evidence for and against the view that a given character 
is ancestral is carefully set forth and only occasionally is it possible 
to decide what Becher's own opinion is.^ The general conclusion 
1 These statements do not apply at all to Becher's later paper ('08) in which 
his discussion of questions of holothurian phylogeny is very clear and satisfac- 
tory. While his conclusions are not wholly in accord with my own it is not 
necessary to discu?s them here. 
