Angiosperrns Tetracentron , Trochodendron , Drimys. 507 
phytes, &c., where there are obvious physiological reasons for the degenera¬ 
tion of tracheary tissue. Furthermore, it is a notable fact that, even among 
such highly specialized plants as these, the complete suppression of vessels 
does not occur in plants having well-developed secondary xylem in their 
roots and stems. Poorly developed or vestigial vessels, easily recognized as 
such by their structural peculiarities, are present in some part of the plant. 
However, there does not appear to be any reliable evidence to indicate 
that Tetracentron , Trochodendron , and Drimys should be placed in this 
category of specialized Dicotyledons. According to E. H. Wilson ( 14 ), the 
well-known botanical explorer, Tetracentron sinense , Oliv., is the second 
largest Dicotyledonous tree in the montane forests of central and western 
China. It attains a height of 16-30 m. and a stem circumference of 4-6 m., 
and has a large crown covered with leaves of extremely herbaceous texture. 
It grows on moist slopes and rich bottom lands in the neighbourhood of 
streams. Trochodendron aralioides , Sieb. et Zucc., like Tetracentron , is a 
large arborescent form which attains a height of 25 m. and a stem circum¬ 
ference of 3-8 m. It grows in moist, warm, montane forests of central and 
southern Japan and of Formosa, and has a large crown with well-developed 
foliage. Unlike these monotypic genera, Drimys comprises a number of 
species which grow in Central and South America, Borneo, New Guinea, 
New Caledonia, Australia, New Zealand, and Tasmania. They are small 
trees or shrubs which usually live in relatively moist environments. Their 
leaves, which vary considerably in size and texture, are not unlike those of 
many Dicotyledons of moist tropical or warm temperate forests. 
In other words, there are no obvious physiological or ecological reasons 
for the degeneration of vessels in Tetracentron , Trochodendron , and Drimys. 
From analogy with the structure of other arborescent and fruticose Dicotyle¬ 
dons, it appears to be highly improbable that such a growth form as 
Tetracentron , with its large crown, thin leaves of herbaceous texture, and 
relatively high rate of transpiration, should once have possessed vessels and 
subsequently lost them, as has been considered to be the case by* Groom ( 2 ) 
and Jeffrey and Cole. 
It has been stated by Jeffrey ( 5 ) that the so-called laws of recapitula¬ 
tion, reversion, ^nd retention are either of universal validity or of little 
scientific value, and that they cannot in certain cases be admitted and in 
others denied. Assuming, for the sake of argument, that such ‘ laws ’ can 
actually be formulated and universally applied in the study of plant evolu¬ 
tion, 1 what significance should be attached to the occurrence of scalariform 
1 It should be noted in this connexion that, even if certain organs of plants are inherently more 
conservative than others, it must frequently be extremely difficult to determine, in the absence of 
reliable collateral evidence, whether a given structure in a given region is cenogenetic or truly 
palingenetic; for even the most ardent advocates of the doctrines of recapitulation, reversion, and 
retention, who tend to minimize or ignore the effects of physiological and ecological factors, admit 
that cenogenetic characters do occur in roots, seedlings, traumatic tissue, &c. 
L 1 
