Bower . — Studies in the Phytogeny of the Filicales . 457 
at or near to the exit of each leaf-trace. The stem of Thyrsopteris is said 
by Diels to attain the thickness of a man’s thigh. In such stems it is 
probable that the vascular system may show greater complexity of polycycly 
than is here described. But this has not yet been seen. 
Saccoloma. 
Of Ferns whose structure has hitherto been figured and described, the 
nearest correspondence to these details of Thyrsopteris is found in a stock 
examined by H. Karsten (‘ Vegetationsorgane der Palmen,’ 1847, IX, 
Figs. 5 and 6), and described under the name of Dicksonia Lindeni. He 
speaks of the Fern as having an upright axis, while it had runners springing 
from the leaf-bases, and their vascular supply came off as a conical or 
cylindrical bulging of the vascular cylinder of the main shoot. This is as 
in Lophosoria and Metaxya , or in Thyrsopteris itself, while the habit appears 
to have been like that of Thyrsopteris , having like it runners which may 
rise upright when strong enough. He found the full-grown stem to be 
polycyclic. The outermost ring was a complete solenostele, as in Thyrso- 
pteris , but in his Fig. 6 it is interrupted at one point close to the outgoing 
leaf-trace. A second ring lay within, more regular and connected than that 
seen in Thyrsopteris , and opposite an outgoing leaf-trace in Fig. 6 a com- 
pensation tongue is clearly shown. 
The Fern named Dicksonia Lindeni by Karsten is now recognized as 
a variety of Saccoloma domingense , (Spr.) Prantl (see Prantl, ‘ Arb. K. Bot. 
Gart. zu Breslau,’ 1893, p. 19; also Christensen, * Index Filicum’). This 
suggests further comparison with the well-known drawings of Mettenius of 
Saccoloma elegans} The details are slightly different, and the stock requires 
a full elucidation ; but the main features appear to be the same. The further 
question, however, arises as to the sorus of Saccoloma. Is its receptacle, like 
those of Dicksonia and Thyrsopteris , of truly marginal origin ? This point 
has been examined on material of Saccoloma elegans collected in Jamaica. 
Unfortunately, owing to the seasonal condition of the plant, it was not 
possible in August, 1909, to obtain all stages of the development, but 
enough was secured to show that the receptacle is of exact marginal origin. 
This is shown by the Figs. 16, a } b. In the first of these (a) a leaf- 
margin is shown with regular segmentation and a marginal cell ; it is, how- 
ever, becoming rounded off prior to the formation of the sorus. Fig. 16, b , 
shows a more advanced state, in which the alternate segmentation is still 
readily recognized, and the marginal cell occupies a median position. But 
the development of the two sides has become unequal, and already the two 
lips of the indusium appear as swellings on either side of the slightly conical 
receptacle, the centre of which is occupied by the marginal cell itself. It 
1 Abhandl. a. d. K. Sachs. Ges. d. Wiss„, vi (1863), p. 531, PI. VI, Figs. 1-4. S. elegans is a 
synonym for S. domingense , Prantl. 
