﻿234 



RELATIONSHIPS. 



Seeds. — The seeds of Cycadeoidea, no larger than a small grain of rye, have 

 many resemblances to and are not, on last analysis, fundamentally different in struc- 

 ture from those of the existing cycads. In the latter there is no distinct demar- 

 cation of the outer flesh, middle stone, and inner flesh constituting the seed [and 

 nucellar] wall into an outer and inner integument. Likewise in Cycadeoidea there 

 is no differentiation into separated integuments. Beneath the outer palisaded layer, 

 as covered in the basal region by cortical cells belonging to the pedicel, lies the 

 fibrous, or in C. Morierei fleshy, hypodermal layer, and interior to this the wall of 

 the uucellus. This double to treble layered testa is so much more reduced than is 

 the testa of modern cycad seeds that, in the absence of a knowledge of other explan- 

 atory variations in fossil forms that will doubtless soon be found, it is still difficult 



to suggest the exact equivalency of the layers. 

 However, in the cycadofilicineau form Lageu- 

 ostoma Loiiiaxi, the seed is also in its general 

 structure cycadean, and there is more resemblance 

 to the cycadeoidean testa. The Lagenostoma cha- 

 lazal region, moreover, as entered by a single 

 concentric bundle, has identically the same radio- 

 symmetric structure as in Bennettites. Likewise 

 (interior to a peculiar cupule) the outer layer of 

 the testa is palisaded. If some such cupule or 

 structure as is seen in Lagenostoma was ever also 

 borne hypogynously in some directly ancestral 

 cycadeoidean seed, its only remnant is the pro- 

 longation of the cortical region of the pedicel, 

 as loose, string}' cells forming no true part of 

 the seed walls. 



The much-elongate micropylar tube of cyca- 

 deoidean forms is a secondary structure correlated 

 with the inclosed position of the seeds between 

 the enlarged tips of the iuterseminal scales. The 

 dicotyledonous embryo is paralleled by that of the 

 existing Cycas, and it is probable that a much 

 varied polycotyledony was present in the cycadeoi- 

 dean as in the existing cycads, there being in this 

 respect no fundamental difference. The apparent absence of endosperm is, however, 

 a far more striking variation, in some way connected with the small size of the 

 cycadeoidean seed. 



The subject of the embryogeny of Cycadeoidea must, however, for the present 

 be left in abeyance, although there is ever}' good reason to hope it will eventually 

 be worked out with reasonable completeness, so far as concerns all larger features. 

 While it is already evident that a more primitive condition is present than in 

 existing cycads, it is scarcely possible, from the facts given in Chapter VI, that 

 structures or modes of growth not primitively and in large degree cycadean are 

 present. 



Fig. 138. — Dictyozamites indicus Feistmantel. 

 Natural size. The netted- veined cycad, 

 illustrating marked diversity in venation 

 of cycadean leaves. (After Feistmantel.) 



