﻿io SCOTT, Early History of Seed-bearing Plants. 



clairi, is of special interest for two reasons, first because 

 each seed is enclosed within a definite lobed cupule, as 

 in the seed of Lyginodendron oldhamium, and secondly 

 because the seeds are borne on a branched rachis, best 

 interpreted as a portion of a fertile frond, probably of the 

 Sphenopteris type, though in the absence of leaflets no 

 definite reference to any special form of sterile frond is 

 possible. 



The other species, L. Kidstoni, appears to have been 

 borne in the same way on the finer ramifications of a 

 naked rachis ; in this case there is no proof that a cupule 

 was present, but the form of the seed itself is beautifully 

 preserved, and shows characters intermediate between 

 those of the petrified species L. Lomaxi and L. physoides. 



The species L. Sinclairi and L. Kidstoni, especially 

 the former, are of importance as indicating how the seeds 

 of the Lyginodendreae were borne. In the case of Lageno- 

 stoma Lomaxi, the seed of Lyginodendron oldhamium y 

 anatomical evidence shows that the pedicel immediately 

 supporting the seed was of a foliar nature ; in L. Sinclairi 

 we see that this foliar organ formed part of the branched 

 rachis of a modified fertile frond. The nearest analogy 

 among recent plants is to be found in the genus Cycas, 

 where the seeds are borne on foliaceous carpels, not 

 aggregated into a cone, but springing directly from the 

 main stem in rosettes alternating with those of the vege- 

 tative leaves. 



So far as present evidence shows, the Lyginodendreae 

 bore fronds of the SpJienopteris type. The stems grouped 

 under the generic name Heterangium have been found in 

 connection, in the lower carboniferous deposits, with the 

 fronds of SpJienopteris elcgans, S. Linkii, and S. dissecta. 

 The fructification is still unknown, but the anatomical 

 characters, which proved a faithful guide in the case of 



