AND AFFINITIES OF STEPHANOSPERMUM. 383 
17X15 ,. Two of them are represented in Pl. 44, fig. 43. Their position in the pollen- 
chamber is indicated by two little crosses close to the rent in the macrospore-wall (text- 
fig. 2, p. 380). Each body is delineated by a delicate membrane whose contour is slightly 
indented or crenate, the convexities of the crenations being directed towards the centre 
of the body. Within this membrane is a slightly carbonised reticulum, which may 
perhaps be regarded as the nucleus. Such types of nuclear preservation are sometimes 
met with in the cells of the female prothallium. Of cilia attached to these bodies there 
is no trace, though the radiate zone or halo of crystallisation occasionally found around 
them is not without some resemblance to a halo of cilia. 
Whilst the size of these bodies supports the view that they may have escaped from 
the cells of the secondary complexes of the pollen-fragments, and their distribution, 
described below, is that which might be expected were they antherozoids, it is evident 
that in their actual preservation they do not closely resemble the antherozoids of any 
known Pteridophyte or Cycad. Traces of the spiral band are lacking. So far as general 
form is concerned, they are Cycadean rather than Filicinean. In their dimensions, 
however, they approach the Ferns. No positive determination is possible till similar 
and better preserved bodies of the kind have been studied in other specimens. The 
supposed antherozoids, oval in form, have a peculiar distribution. Three occur in the 
mucilage above the ruptured spot, and others at the spots indicated by small crosses in 
text-fig. 2. Of their distribution, it may be said that the antherozoids are met with 
either in the neighbourhood of those spots at which it is reasonable to suppose (in view 
of the other evidence) that antherozoids were liberated, or in the neighbourhood of the 
female prothallium, and nowhere else. Their dimensions harmonise with those of the 
secondary cell-complexes of the pollen-ceils. 
Though there seems little reason to doubt that the bodies in question represent nuclei 
enclosed in loose crenate envelopes, it must be borne in mind that the pollen-chamber 
itself has, no doubt, become excavated by a lysigenous breakdown of the tissues, and it ` 
would be difficult to prove that our nuclei may not have originated in that way. Or, 
again, it is possible they may be the free spores of one of the Fungi known to have 
attacked the tissues of the nucellus in more than one of the Palaeozoic seeds. The most 
striking facts about them are their size and distribution. The attribution of these 
structures to antherozoids is conjectural, but the facts seem to justify allusion to them, 
Even if this seed were recent and not fossil, their occurrence would need further study 
and confirmation. In the present case, where so many pitfalls beset the investigator, 
such corroboration is, of course, more urgent. 
The Question of Pollen-tubes. 
If we look at the question from the wider standpoint of all the facts that have been so 
far elicited, how does the matter stand? In the case of the pollen of Ætheotesta we 
have Renault’s interesting observation of a funnel-like process at one pole of the grain *. 
In Codonospermum oliveforme the pollen-grains (stated to be derived from Dolero- 
phyllum) are provided with an operculum. In both cases the indication would seem to 
* «Sur quelques Cryptogames hétérosporées, p. 7, and fig. 13, p. 5. T Renault, loc. cit. p. 7, and fig. 14, p. 5. 
