SPHSEROSTOMA OVALE FROM PETTYCUR, FIFESHIRE. 13 



of Heterangium Grievii in the same block with the ovules here described. Many of 

 the plants are exceptionally well developed, and three are giving off adventitious roots. 

 In the course of three years, during which time I have had hundreds of slides from 

 this block under observation, I have not been able to detect a trace of any other 

 plant the nature of whose sporangia is not known, except two undescribed stems 

 apparently belonging to immature plants or other species of Heterangium. 



Hence the fact that numerous ovules lie among the petioles and rooted stems of 

 Heterangium Grievii is undoubtedly suggestive of actual reference. 



The argument from association is supplemented by internal evidence. The ovule 

 shows an unmistakable resemblance to Lage?wstoma, which is acknowledged to be 

 the type of ovule borne by Lyginodendron. 



But Lyginodendron, both in the external morphology of its vegetative organs and 

 in their anatomy has long been regarded as showing close affinity with Heterangium. 

 The resemblance between their megasporangial members would be similarly great if 

 Sphserostoma ovale belonged to Heterangium. Such differences as do obtain between 

 the two ovules are such as would be expected between an older and a later type. 

 In Sphserostoma the canopy shows less complete integration (i.e. the crests are free) 

 and the mechanical tissue is less developed than in Lagenostoma. The method of 

 dehiscence of the nucellus is more fern-like and the pollen grains are smaller. These 

 differences are all consistent with the view that we are dealing with a more primitive 

 ovule in Sphserostoma than in Lagenostoma, but with one in which the ground-plan 

 of the Lyginodendron ovule had been laid down. 



This evidence from comparative considerations receives important corroboration 

 from certain features of Dr Scott's slide, C.N., 387, already referred to. 



Many of the ovules are still surrounded with an intact cupule and hence, if the 

 interpretation of the tissue in the pedicel as an abscission layer be correct, must have 

 been still attached to the frond, or part of the plant on which they grew. 



To investigate this matter the three longitudinal sections which were nearest the 

 median plane were examined, with the following results : — 



a was found to be exactly radial at its base ; 



b was radial at the micropyle but slightly tangential at the base ; 



c was in the plane which passed through the plinth but just escaped the pollen 

 chamber. At the base it also just escapes the pedicel of the ovule. 



Sections a, b, and c were found in each case to be accompanied by a petiole in the 

 expected position and cut in a plane exactly corresponding to that of the ovule. Thus 

 a shows a radially cut petiole, b shows one in which the vascular bundle is just touched 

 upon, and c shows one with only the cortex occurring. For the case referred to as 

 a, reference should be made to PL II. fig. 9. 



As already pointed out in an earlier part of the paper, most of the ovules show no 

 cupule. Such ovules in no single case, although they are far more frequently met with, 

 show similarly disposed petioles of Heterangium. 



