Erythrococca and Micrococca. 581 
does not apply to the action of Thwaites in 1861 or to that of Baillon in 
1862, because both of these authors have fully recognized the real affinity of 
Bentham’s genus. Guided, however, by somewhat different considera¬ 
tions, they have arrived at different conclusions. There is but one species 
in Ceylon with which Micrococca Mercurialis may readily be compared ; 
the field in which Thwaites had to work was therefore a very restricted one. 
Knowing that this other Ceylon species had been identified by Baillon 
with a plant from Java which is undoubtedly a Claoxylon , and finding that 
as regards their stamens, interstaminal glands, and hypogynous scales this 
plant and Micrococca Mercurialis entirely agree, Thwaites was left with 
no option in the matter. His decision to reduce Micrococca to Claoxylon 
was probably strengthened by the fact that he had, through some inadver¬ 
tence, formed the impression that Micrococca Mercurialis is a dioecious 
plant; the fact that it is an annual one has appeared to him too unimportant 
for mention. Baillon, on the other hand, as we have already seen, satisfied 
that a striking difference as regards the position of their anther cells is 
of no real consequence, reverted to the conclusion of Lamarck, and replaced 
Micrococca in Mercurialis , Linn. 
Mueller in 1866 mentioned, but could not accept (DC. Prodr., xv. 2, 775) 
the conclusion of Baillon ; that of Thwaites he adopted in 1865 (Linnaea, 
xxxiv. 166), notwithstanding his belief that the stamens of Micrococca differ 
as regards their dehiscence from those of Claoxylon} But the action of 
Mueller is less natural than that of Thwaites. In addition to making use 
of the staminal character in question, Mueller noted that in Micrococca 
Mercurialis male and female flowers may occur on the same rachis, so that 
the plant is not dioecious as Thwaites has stated ; he further laid stress on 
the fact, not alluded to by Thwaites, that Micrococca Mercurialis is an 
annual. Guided by these considerations, Mueller kept the two plants which 
form the genus Claoxylon , as understood by Thwaites, rather widely apart : 
Micrococca Mercurialis he treated as the type of a distinct section ; the 
other, which he now recognized as perfectly distinct from the Java plant 
(Erythrochilus longifolius , Bl.) with which it had been identified by Baillon, 
he nevertheless placed in his own section Euclaoxylon, with which, except 
for possessing interstaminal glands in the male flower, it has wonderfully 
little in common. It is hardly surprising, therefore, that in 1880, Bentham 
(Gen. Plant., iii. 309), while he agreed with Mueller in regard to Baillon’s 
conclusion, should have felt unprepared to accept the view, either as stated 
by Thwaites or as modified by Mueller, that the genus Micrococca should 
be merged in Claoxylon . 
Yet when we examine the arguments employed by Bentham in favour 
of the maintenance of Micrococca we find them as little convincing as those 
1 This belief turns out to be without foundation; the anther cells of Micrococca open precisely 
as those of Claoxylon proper do. 
