450 



OPHIOGLOSSALES 



were of irregular construction from the first, as above described. It is 

 thus seen that in O. pendulu?n only a comparatively small residuum of the 

 original sporangiogenic band finally remains as fertile tissue. Meanwhile, in 

 the broad bands of sterile tissue which have thus been initiated between 

 the sporangia, vascular bundles make their appearance, connected as 

 branch-bundles with the general system of the fertile spike (v. B, Fig. 249 

 D, and Fig. 250) : in those cases where the arrangement is regular they 

 may occupy a definite position, corresponding very nearly to the point of 



FIG. 249. 



Ophioglossum (Ophiodenna) pendulum, L. A, .5 transverse sections of spikes of 

 different ages to show sporangia and vascular bundles, slightly enlarged. C = a single 

 sporangium, older than in Fig. 248, seen in longitudinal radial section ; the tapetum lightly 

 shaded surrounds the darker sporogenous mass. D = tangential section of corresponding 

 age, showing one sporogenous mass shaded ; the smaller shaded groups will form the 

 vascular strands. -"=part of a transverse section of an older sporangium of O. reticu- 

 latum. X loo. 



intersection of the lines limiting the cell-groups of the original sporangio- 

 genic band. In certain cases, where the segmentation is regular it appears 

 that one sporangium is referable in origin to two of those cell-groups, but 

 it cannot be said that it is always so : frequently the arrangement is 

 irregular, and in any case the single sporangium cannot readily be referred 

 in origin to a single parent-cell. 



Examination of the young spikes of O. vulgatum and reticulatnm 

 shows that in all essentials the development is the same, though naturally 

 with differences of detail and proportion in those less bulky species. In 

 them it is also impossible to refer each sporangium to a single parent cell. 

 Further, it has been shown in them that the archesporium is not hypo- 

 dermal in the strict sense, that is, that it is not cut off once for all by 

 one periclinal wall or walls, but that successive additions may be made 



