144 BOTANICAL GAZETTE [februasy 



size, which is in toto much greater than that attained by the two 



be such 



cotyledons of a cycad. 

 in the pine seed the production of the massive endosperm and indurated 

 coats that characterize the cycad seed ; and let the cotyledons be con- 

 fined within the moist endosperm until they attain their full size, 



them 



nection with it as to cause difficulty in distinguishing between them, 

 and bringing the tips of the cotyledonary vascular strands into intimate 



■ _ * m _ 



>erm 



Cycas revoluta (13) and by Thiessen for Dloon edule (12): under 

 such pressure, the inner faces of the pine cotyledons would be very 

 intimately united and the question naturally suggests itself, What 

 would become of the epidermis of these inner faces ? 



The alternation of mucilage canals with the cotyledonary vascular 

 strands in cycads, and its ready relation to the peculiar condition 

 found in pine cotyledons, may be used as evidence for a theory of 

 fusion as well as for one of splitting, as Hill proposes (8) . 



The absence of the protostele in the hypocotyl of Microcycas in 

 contrast to the condition found in Dioon edule and Ceratozamia may 

 not have any significance in the light of recent investigation. That 

 the protostele is, in general, the most primitive condition of the vascular 

 axis may be true; but that this structure must occur in every primi- 

 tive vascular plant is, of course, not true; neither are we to regard a^ 

 primitive all plants in which it is found. Chrysler (4) has found it 

 in members of the Araceae. There are other characters, however, 

 which seem to indicate a greater advance than that made by Cycas or 

 bncephalartos, or even Ceratozamia. These are the single stele and 

 the degree of elimination of the cortical cambium, which, in the cycad 

 >tems, produces this vascular tissue. However, the large proportion of 

 centripetal wood in the foliar traces is an offsetting primitive character, 

 which must be weighed in the same balance. 



1 he undoubted polyspermy would seem, at first sight, to bear 

 down the weight of evidence on this side; but it is possible that this 

 primitive feature is a recurrence rather than a direct inheritance: 



I , jE T FFREY calls a coenogenetic, rather than a palingenetic, char- 

 ier. j UEL (Q) found ag ^ ^ erms ^ the poUen tub e 



ot Lupressus Goveninnn ™,i , . ., , ^..^n^ns has 



