CONIFERALES (TAXACEAE) 



349 



drical mass of meristematic cells, and this cylinder invades the endo- 

 sperm, its basal cells contributing to the suspensor. Later, growing 

 points are organized, the root tip being deep in the cylinder, and the 

 two concave cotyledons completely surrounding the stem tip. The 

 peculiarity of the embryo-formation in 

 Cephalotaxus (25, 124, 130) has been 

 m.entioned, in which the terminal cells of 

 the proembryo form a persistent cap for 

 the embryo during its advance through the 

 endosperm. That this same feature appears 

 in Araucaria brasiliana (25) is a matter of 

 interest (figs. 367, 368). 



So far as known, all of the Taxaceae 

 are dicotyledonous. This has been re- 

 garded as a primitive feature, because 

 Bennettitales, Cycadales, and Ginkgoales 

 are dicotyledonous; but the relation of 

 polycotyledony to dicotyledony may lead 

 to a different conclusion (see p. 298). 



Podocarpus Makoyi has been found 

 (73) to be viviparous, the embryo con- 

 tinuing its development and the hypocotyl 

 passing out of the micropyle while the 

 seed is attached to the parent plant. It 

 is probable that other species of Taxaceae 

 will be found to show the same habit, 

 as it is not uncommon in other groups of 

 gymnosperms. 



5. History and distribution 



The Taxaceae have not been recognized farther back than the 

 Cretaceous, and so far as this negative evidence can be rehed 

 upon the family is comparatively a modern one among gymnosperms, 

 certainly much younger than the Abietineae and the Araucarineae. 

 It is possible that older Taxaceae existed, and that they have not been 

 recognized among coniferous remains; but until they are recognized, 

 the failure to find them earlier than the Cretaceous must be considered 

 in any discussion of phylogeny. 



Fig. 403. — Torreya taxi- 

 folia: proembryo showing 

 elongation of suspensor cells; 

 April 7, 1904; X460. — After 

 CoxiLTER and Land (ioi). 



