576 Prain.—A Review of the Genera 
Baillon would include in his widened Mercurialis the anther cells at first 
have an erect position. This character, superadded to those others which 
Baillon admits to be of sectional value, makes it as easy as it is convenient 
to treat Mercurialis , Linn., as a distinct genus. But if an intrinsic staminal 
character may be used for one genus we can legitimately employ such a 
character in connexion with another. In this way the separation of Seidelia , 
Baill., is also simple, for in Seidelia the anther cells open longitudinally both 
in front and behind so that, when their pollen is shed, the anthers are 
cruciately 4-valved, whereas in Adenocline, Micrococca , and Claoxylon the 
anther cells open, as Baillon states that they do in Erythrococca , along one 
side only, ‘ par une fente longitudinale dont les bords s’ecartent beaucoup 
et se reflechissent en dehors, de maniere a donner a la loge ouverte la forme 
d’un cornet ’ (Etud. gen. Euphorb., 437). A criterion from the same source 
might be used to distinguish the genus Adenocline , Turcz., because in 
Adenocline the empty anther cells assume the downward position character¬ 
istic of those of Mercurialis while still in bud. In this instance, however, 
the need to rely on a staminal character does not arise ; Adenocline is 
readily separable from the other types under discussion, because the calyx 
in the male flower is not closed in bud as it is in Mercurialis , Seidelia , 
Mirococca , Erythrococca , and Claoxylon. At this point, however, staminal 
characters cease to be effective because the anthers of Micrococca , Claoxylon, 
and Erythrococca cannot be distinguished at any stage. 1 Had staminal 
characters alone been legitimate in the discrimination of the genera with 
which Baillon deals, the reduction of both Micrococca and Erythrococca to 
the older genus Claoxylon could be effected without further discussion. In 
the case of Erythrococca , no one has hitherto suggested that such reduction 
is necessary ; in that of MicrococcaW. has been effected by Thwaites (Enum. 
PI. Zeyl., 271), whose action has been accepted by Mueller (DC. Prodr., xv. 2, 
789) and confirmed by Hooker (Flor. Brit. Ind., v. 412). 
But even if we should grant that it is necessary to merge both Erythro¬ 
cocca , Benth., and Micrococca , Benth., in Claoxylon , A. Juss., this would fall 
short of accepting the position which Baillon postulates. What Baillon asks 
us to admit is that in deciding that Micrococca is congeneric with Claoxylon 
and Erythrococca we must be guided solely by the fact that as regards their 
stamens the three are indistinguishable ; whereas, in deciding that Micro¬ 
cocca is congeneric with Seidelia , Adenocline, and Mercurialis we must treat 
their staminal differences as negligible. This position has not been accepted 
by Mueller (DC. Prodr., xv. 2, 775) or by Bentham (Gen. Plant., iii. 309); it is 
1 Mueller (DC. Prodr., xv. 2, 789) has stated that in Micrococca the anther cells open introrsely, 
whereas in. Erythrococca ( 1 . c., 790) and in the sections Adenoclaoxylon, Athroandra, Gymnoclaoxylon, 
and Euclaoxylon of Claoxylon ( 1 . c., 775, 776, 780, 781) they open extrorsely : he is silent as regards 
Discoclaoxylon, where also they open extrorsely. But, as Bentham has already pointed out (Gen. 
Plant., iii. 309), there is here some error in observation. An examination of the anthers of Micro¬ 
cocca shows that the cells do not open introrsely. 
