1909] THOMSON— SAXECOTHAEA AND MICROCACHRYS 351 



(fig. 21). In Jig 



norma 



— bast below 



and wood above, not shown in the section— while cutting obliquely 

 across the field is a sporangial supply bundle with bast above and 

 wood below. A short ventral extension to the sporangia turns at 

 right angles at first and then becomes almost inverse. Figs. 22 and 

 I are transverse sections of a sporophyll. In the upper part of the 

 figures are the main bundles, with wood above and bast belw. 

 Farther down and toward the ends of the normal series are some 

 inversely oriented bundles. One, to the left, magnified in jig. 23, 

 has just divided into a similarly oriented bundle, and from this 

 there is a strand to the sporangium. The latter has a more or less 

 concentric character. In some cases I- have found a second inversion 



on the way to the sporangia. 



The inversion of microsporangial supply bundles is of wide occur- 

 rence in the cycads, but is not confined to the group, being also 

 found in certain conifers. This feature has not been considered 

 by the exponents of the theories advanced to explain the inversion 

 in the fertile scale of the latter. That it is of first importance seems 

 self-evident. Again, the double inversion in the fertile scale has not 



been explained, and a glance at the figures shows how important the 



... ' e .... :„~„ wMth tho simile 



*cond 



Microcachrvs 



-"viwuii in euner ^axegomaea. «>i jbiuw*«»;- u f the 



'?)• This and the occurrence of ovules on the lower surface 01 1 



ale in Saxegothaea, and their vascularization from the main supp^ 

 *ould seem to be good reason to consider that inverse ^^ 

 ^nnot in all cases be relied upon as evidence of the braci} 

 character of the fertile scale. In the last case, we should nav 

 anomaly of the leaf in the axil of the shoot. 



Again, in Dacrvdium two bundles with orientation *"*» 

 ? the scale supply the ovule. These attach M » *£* 

 e e passing close together near the upper surtace 

 ere is, then, in Dacrydiu-n, a form* ■£»**' *J5 



thre 

 Th 



Microcachrj 



-•■ «ma oaxegoinaea or ivuuu^a^uij^ more 



the cone, etc., a greater amount of inversion than 1 ^ 



Primitive forms-a puzzling feature certainly in vie 

 br achyblast theory, but possibly finding its explanation in , 



