258 Hill and de Fraine . — A Consideration of the 
what remarkable unanimity of opinion, for it appears to be generally 
fashionable to consider the more complicated structures to be the more 
primitive, and from them to derive most simpler organizations. Indeed, 
there appears to be some danger of the theorem ‘ Reduction in Descent ’ 
becoming an axiom . 1 
This is perhaps natural, for it is certainly easier to work on this 
hypothesis than on its reverse : it is pretty safe to prophesy that Macaulay’s 
New Zealander, when he does come to England and rummages in the ruins 
of Science Museums, will conclude that Stephenson’s ‘ Rocket ’ and similar 
engines are reduced structures, and that the primitive locomotive is to 
be found in some such form as the ‘ Duchess of Cornwall 5 ; he will certainly 
recognize in the ‘ Atlantic ’ a synthetic type, and he will indeed be a genius 
if he sees in the tea-kettle the forbear of the ‘ Rocket ’. 
We do not, of course, desire to imply that reduction in descent does not 
obtain, but rather to express the opinion that there is sometimes a tendency 
to invoke it without due consideration of the alternative proposition. 
Either proposition is merely a generalized statement ; no doubt if the 
ancestry of any plant were intimately known, it would be found that its 
evolutionary curve is neither / nor \ , but either /sfS' or ; that is 
to say, its history would show periods of specialization and amplification 
and periods of reduction, the general character of the curve depending on 
which side the resultant falls. 
The matter is further complicated by the fact that a plant is a complex 
of characters, some of which are primitive, others reduced — sometimes to 
a condition which is indistinguishable from the primitive — whilst others are 
specialized, and, unfortunately, practically nothing is known about the inter- 
action between the different characters. Thus in the case of Gymnosperms 
it does not follow, we think, that ‘ the coniferous forms show reduction 5 — 
with regard to their seedling anatomy — £ on the Cycad type, ... on the 
grounds of the generally acknowledged antiquity and primitiveness of the 
latter group ’. 2 
And this opinion, apart from other considerations, which will be dealt 
with later, is supported by the fact that the cotyledons of Ginkgo and some 
Cycads are characterized by the presence of stomata . 3 This is a very 
surprising fact if the seed leaves of these plants have always been hypogeal 
and embedded in the endosperm of the seeds ; it rather indicates that the 
hypogeal habit is a derived habit. Further, it may be remarked that neither 
tetrarchy nor diarchy is dominant in these plants , 4 and also that there 
1 Tansley, A. G. : Reduction in Descent. New Phyt., i, 1902. 
2 Thomas, E. N. : A Theory of the Double Leaf-trace founded on Seedling Structure. New 
Phyt., vi, 1907. 
3 Wigglesworth, G. : The Cotyledons of Ginkgo biloba and Cycas revoluta. Ann. Bot., xvii, 1903. 
4 Hill, T. G., and de Fraine, E. : Gymnosperms, Pt. Ill, loc. cit. 
