The Old IVood and the New. 
29 
In the series we have followed, we have seen how the centrifugal 
wood, enormously reinforced by secondary growth, has gradually 
exterminated the old and unprogressive centripetal wood, which 
becomes more and more a mere vestigial structure, until it is lost 
altogether. The Cryptogamic wood was, in fact, doomed, from the 
moment when secondary growth, with its unlimited possibilities, 
established itself. IT has only survived as long as it has, because 
of the conservatism of the foliar bundles. In the leaf there was 
not much demand for secondary tissue, and so the old wood was 
long able to hold its own—down to the present day in the Cycads. 
In the leaves of Conifers, too, as Mr. Worsdell has given us good 
grounds for believing, the centripetal wood still persists in the form 
of transfusion-tissue. It has survived here by the help of a change 
of function, similar to that, which, as I have suggested, enabled 
the central wood in the stem of Megciloxylon to justify its existence, 
after it had become superfluous as a conducting tissue. 
For a time the leaf-traces connected with the foliar bundles 
retained the old wood, and they retained it longest in the part of 
their course nearest the leaf. 
In the peduncles of some Cycads, organs of limited duration 
with little need for secondary growth, the old centripetal wood has 
lingered on, in a sadly reduced form, to the present day, but can 
scarcely now be of much use to the plant—“ superfluous lags the 
veteran on the stage!” 
£ 
These are the only stem-structures in Phanerogams which 
still retain the Cryptogamic wood, if we except the curious case of 
Rothert’s Cephalotaxus Koraiana , where centripetal wood (not 
secondary in origin) seems to have re-appeared in great force; 
so isolated a case can scarcely be regarded as a survival. 
The series which w r e have followed—from Heterangiam to 
the Cordaiteae and Cycads—is not a chronological one, and makes 
no claim to represent the true course of descent. It is only, 
however, among Palaeozoic plants that our transitional forms 
occur; by the time we reach the Mesozoic Gymnosperms, the old 
primary wood has, so far as we know, disappeared from the 
stem. Further, there is no doubt that the extinction of the centri¬ 
petal wood coincides, on the whole, with the general replacement 
of Cryptogamic by Phanerogamic characters. By the use of such 
provisional series as that which we have sketched in the preceding 
