Internodal Vascular Strands of Equisetum. 21 
been recorded from the characteristic palaeozoic Equisetales, the 
Calamariae. It is true that a few years ago Mr. Kidston (10) figured 
and gave a short description of what he termed a “ rhizomatic 
tuber,” belonging to an undetermined Calamite of Westphalian age. 
In this case, however, the tuber consists in three swollen internodes 
of a rhizome, the other internodes of which appear to be unaffected. 
This certainly seems a less specialized condition than the develop¬ 
ment of “ chaplets” of tubers or of special branches as single tubers. 
It should noted that Duval-Jouve figures a rhizome of E. arvense 
in which a single internode is swollen and tuberous (6, PI. I, Fig. 2). 
Again, the type of bundle in which the lateral metaxylem is 
absent, or reduced to a cell, seems clearly to have arisen by 
reduction of the metaxylem from the bundle with three separate 
groups of xylem. Between this last-mentioned type of bundle and 
that with continuous xylem there exists, as has been shown above, 
a fairly complete series of intermediate forms. We can hardly 
doubt that one of these types has arisen from the other. Which, 
then, is the more primitive ? The question is one of some 
importance, for, if the three groups of xylem have arisen in the 
phylogeny by the breaking up of a single mass of xylem, owing to 
the failure of certain cells to develop as tracheides, then the whole 
bundle is, as has been recently claimed by Eames (7, pp. 593-524), 
clearly a phylogenetic unit. 
It seems to me that the arguments in favour of the priority of 
the bundle with continuous xylem are very strong. In the first 
place such priority would assume the gradual reduction of the 
bundle in the Equisetaceae, and few, if any botanists, would now 
dispute the proposition that the existing species of Equisetum are 
reduced forms. There can be little question but that some of the 
Mesozoic Equisetites were among the ancestors of the recent genus 
Equisetum ; and these mesozoic fossils show, on the whole, 
a diminution in size as we pass up from the older to the more 
recent rocks. Secondly, it is clear that in the rhizome the vascular 
tissues are, on the whole, markedly more reduced and it is 
here that we find hardly any indications of the continuity of 
metaxylem and protoxylem. I certainly regard it as significant 
that in the only fossil rhizome of which we possess a detailed 
anatomical account the lateral metaxylem is more highly developed 
than in the recent forms and is continuous with the protoxylem. 
This rhizome, known as Equisetites noviodunense , comes from the 
Sparnacien, and it would seem that at this geological horizon the 
reduction of the vascular tissues, even in the rhizome, had not 
