XXXIII] GAMETOPHYTE 303 



In Fig. 566 the stalk of one of the sporangia has been traversed, showing 

 this structure in the young state, and it persists without further division till 

 the sporangium is ripe. This type of cleavage has been verified for all the 

 three genera of the Cyatheaceae, and is found to be a constant feature for 

 them. It has already been described for Metaxya, and it will be seen to exist 

 also in Dipteris and Cheiropleiiria. It thus appears to be widespread among the 

 Superficiales. No instance of it has yet been seen among the Marginales. 



If the problem of segmentation in a phyletic sequence with diminishing sporangia be 

 considered, it appears that more than one course is open, just as it is seen to be in the 

 segmentation of a conical apex of stem or leaf or root, where the segmentation may be by 

 four, three, or two rows of segments. A reference to Fig. 243, Vol. I, jj. 248 presents the 

 problem in terms of the sporangial stalk, where it is naturally simpler than it would be in 

 the sporangial head itself A four-sided segmentation has actually been seen in certain 

 sporangia of Todea, and the Osmundaceous stalk may perhaps reflect that structure with 

 four internal cells (Vol. i. Fig. 243, d). As the sporangia diminish the internal cells may 

 be omitted, so that the cleavages of the initial segmentation become more apparent in the 

 simplified stalk. If the type of the sporangium were one with a relatively short thick stalk, 

 as it is in the gradate Superficiales, it might well be expected that an alternate cleavage 

 would suffice for the formation of the relatively small head, while subdivision of the two 

 rows of segments would readily yield a massive four-rowed stalk. On the other hand, if 

 the sporangial stalk were elongated and thin, and the sporangial head proportionately 

 larger, then a three-rowed cleavage would provide the sporangial head, and would give 

 without sub-division the three-rowed stalk, which is the rule for most Leptosporangiate 

 Ferns, excepting those that are the most advanced. Accordingly it may be held that the 

 short thick stalk and diminished head of the sporangia in the Cyatheoid Ferns, and their 

 kin, may have determined a segmentation that is at least sufficiently unusual to have 

 escaped observation hitherto. 



The Gametophyte 



The gametophyte of the Cyatheaceae presents no special features when 

 grown normally. It is of the usual cordate t}'pe. But in old prothalli certain 

 characteristic bristles are produced on both sides of the prothallus, a feature 

 which recurs in other allied Leptosporangiates such as Diacalpe and Woodsia 

 (compare above, p. 267: also von Goebel, Organographie, ii, p. 950). 

 Prof A. S. Stokey permits me to state that she finds in pure cultures of 

 AlsopJiila and other Cyatheaceae small flattened scales comparable to those 

 of the sporophyte. They appear usually after the archegonia on old prothalli. 

 The chief interest of the gametoph}'te is, however, centred in the antheridia. 

 Their early segmentations follow those usual in Leptosporangiate Ferns, but 

 with the significant difference that after the distal cap-cell has been formed it 

 may divide further, as has been seen also in Diacalpe and Woodsia, genera 

 closely associated with the Cyatheaceae (see Heim, 5 56). The number of sper-' 

 matocytes traversed in a median vertical section of an antheridium of Cyathea 

 medidlaris is seen from Bauke's drawings to be 15-18 {I.e. PI. viii, Figs, i, 8). 



