36 " ENDEAVOUR " SCIENTIFIC RESULTS. 



corbnla and curving forward over it, and having its edges fringed 

 with sareothecae. The lateral spur-like projections from the 

 distal edges of the corbula-ribs near the base are large and pro- 

 minent, and are armed with eight or ten, or even more, 

 sarcotheca\ In the Museum specimens, however, I found 

 narrower corbulee, with the lateral spurs reduced to short 

 stamps carrying only three or four sareothecae, while the free 

 borders of the leaves were very narrow and the superior foliaceous 

 extensions entirely wanting, or only very slightly developed. In 

 some specimens fortunately the gonophores were sufficiently 

 preserved to enable the sex to be recognized, and in all such 

 cases I found that the corbulae with the large secondary leaves 

 contained female gonophores in various stages of development, 

 and iu some cases planulae ready to emerge, while all those of 

 the otlier type contained male sporosaos. 



I have no doubt that a similar differentiation will be found in 

 the other species nearly allied to A. billardi. In A. tasmanica 

 I have found the female gonophores present in corbulae which 

 even exceed those of A. billardi in the extent to which the 

 secondary structures are developed, but I have not yet seen the 

 male corbulae. Ritchie mentions that in his specimens of 

 A. -cruciali s "the margins of the leaflets frequently rise into 

 very pronounced crests." In A. macrocarpa I have seen only 

 one specimen with corbulae, which, from analogy with those of 

 A. billardi, are probably male. It may be noted too that in 

 these corbulae the distal portions of the leaves do not quite meet 

 in parts, so that a number of small openings are left between 

 them, as in some of the forms mentioned by Torrey. 

 A. clannevigi lias corbulae which differ among themselves in a 

 fashion corresponding to that of A. billardi, though not in such 

 ;i marked degree, and in this case also the difference is presumably 

 sexual. In another (undescribed) species, not at all closely 

 related to any of the foregoing, the sexes are easily distinguish- 

 able, and here also the female corbulae are marked by the far 

 greater development of the structures in question. It is evident 

 that none of the species which possess these appendages can be 

 considered sufficiently known until both sexes have been 

 identified. 



In A. billardi (as also in A. tasmanica) the substance of the 

 corbnla-leaves, like that of the hydrothecae, is nearly or quite 

 colourless, and the crests have a somewhat flaccid appearance, 

 being often bent about irregularly, yet the perisarc is really very 

 thick, or rather it consists of two laminae forming a hollow wall, 

 and separated by a considei'able interval. The corbnla reaches 

 a great length, one or two specimens which I observed measur- 



