696 A TEXTBOOK OF THEORETICAL BOTANY 



recognized to be more nearly related to the Araucariineae and included in 

 the Coniferales. 



Taxus has a single species which is distributed over Europe and the 

 Mediterranean region, as far east as Persia. Its northern limit is about 

 61° N. It also occurs in eastern Asia and in North America. 



Torreya is limited to one species in Japan, one in California and one in 

 Florida. It is sometimes placed in a separate family. 



Cephalotaxiis has five species in south-east Asia and one in southern 

 Japan (Fig. 699). 



They are separated from the Coniferales by the absence of a female 

 cone, the ovules being borne singly at the apices of very short shoots and on 

 different plants from the males. In Cephalotaxiis the ovuliferous shoot has 

 been reduced to the vanishing point, so that the ovule appears to stand in the 

 axil of a bract. 



The nucellus is completely free from the integument in its early stages, 

 but during development after pollination there is a considerable amount of 

 intercalary growth at the base or chalaza of the ovule, so that eventually the 

 free portion of the nucellus represents only a small apical part of the whole. 

 Normally only a single seed matures from each female inflorescence and 

 this is surrounded by a fleshy outer covering, the so-called aril, which may 

 be free from the integument or united to it, and is considered by some to be 

 equivalent to a second integument. 



The pollen grains are smooth and wingless. No prothallial cells are 

 formed in the male gametophyte and in Taxus and Torreya the cytoplasm 

 of the male cell enters the oosphere with the fertilizing male nucleus and 

 envelops the fusion nucleus. In the archegonium there is no evidence of any 

 ventral canal cell being formed, though a very ephemeral ventral canal nucleus 

 may be produced, which is evidently on the way to complete elimination. 

 This is apparently also the case with the megaspore membrane, which is 

 scarcely developed at all and never thickens as it does in the Abietineae. 



Torreya has features in the seed which are highly peculiar. There is a 

 vascular plate at the base of the nucellus which sends branches into the 

 nucellus itself and these connect with a peculiar mucilaginous layer surround- 

 ing the nucellus, which, it has been suggested, may represent the vascular 

 mantle that encloses the nucellus in some Palaeozoic Pteridosperm seeds. 

 Nucellar vascular tissue is found in the relatively primitive Cycadales, but it 

 has disappeared from the Coniferales. The basal growth of the ovule carries 

 the nucellar bundles upwards, so that in the mature seed they are found 

 entering the nucellus at the bottom of the apical, free part, through openings 

 or foramina in the stony wall of the testa. Torreya also has a very vigorously 

 growing endosperm, which encroaches on the surrounding tissues in an 

 irregular fashion, producing an extremely complicated outline. This is also 

 found in a few Angiosperms and is called ruminate endosperm. 



There are certain points of vegetative anatomy which distinguish the group. 

 One is the presence of centripetal xylem in the cotyledonary bundles, 

 occasionally also in the stem axis and, notably, in the aril of Cephalotaxus, 



