743 
Seedling Anatomy of Certain Sympetalae . 
The great majority of the species comprised in the Tubiflorae are 
herbaceous, and in many respects this extensive group is generally regarded 
as a remarkably natural one. Engler and (especially) Bentham and Hooker 
recognized certain sub-groups ; and in the latest attempt to deal with the 
Tubiflorae, Wernham (29), discussing the question from the point of view of 
floral evolution, appears to hold that certain of these sub-groups are not only 
well defined, but possibly had a quite different course of evolution. In this 
connexion, it is perhaps worthy of note that in the principal sub-groups, 
i. e. in Polemoniales, Personates, and Lamiales of Bentham and Hooker, the 
prevailing method of transition is Type 3, while in each of the groups 
Polemoniales and Personales a single Natural Order shows (in part) 
a different arrangement. Although it is not possible to advance any 
external cause as being responsible for this difference in the seedling 
anatomy, the cases quoted serve to emphasize the fact that orders, and even 
genera (e. g. of Bignoniaceae) which are generally regarded as being closely 
allied, show different methods of transition. Conversely, just as de Fraine 
demonstrated in the delimiting of Cactaceous genera , so here with regard to 
the sub-groups and (rarely) orders of the Tubiflorae, no assistance is 
derived from seedling anatomy. 
Perhaps the chief interest which emanates from the present study 
of the seedling structure 01 the Tubiflorae attaches to the occurrence of the 
Anemarrhena type of transition. Previous to the discovery of this type in 
the Cactaceae (6), it had been found only in certain Ranunculaceae among 
dicotyledons (13, 17), and on its occurrence here, as well as on other 
features common to this order and to certain monocotyledons, had been 
founded a theory of the origin of the latter group from a dicotyledonous 
ancestor. In opposing this view to the light of her discoveries in the 
Cactaceous genera, Opuntia and Nopalia , de PTaine concludes (6, p. 167): 
‘ It cannot be considered that the resemblance of Opuntia, for example, 
to Anemarrhena is the result of a close genetic relation between the two ; 
nor can it be conceded that it is due to the response of two unrelated forms 
to similar conditions ; hence we cannot but conclude that the resemblance 
is accidental. This being so, then it is quite possible that the similarity 
of Era7ithis and Podophyllum to Anemarrhena is also accidental.’ The 
conclusion arrived at is that ‘ the evidence on which the “fusion ” hypothesis 
[of Sargant (14)] was based has been considerably weakened by the dis- 
covery of the Anemarrhena type of seedling in such a specialized order as 
the Cactaceae ’. 
This conclusion, however, is not generally admitted. It has been 
called into question by (Mrs.) A. Arber (1) on the ground that the Cacta- 
ceae, while showing a high degree of specialization in their vegetative 
characters, may really be closely allied to the Ranales. A comparison was 
made between the floral structures in the two orders, Nymphaeaceae and 
