PARKA DECIPIENS. 7 



Myretou collections. They differ, however, entirely in structure from Pachylheca, which 

 is found in the same beds, but retaining its rotundity of form (while Parka is flattened), 

 and showing no cellular areolation. Nor is there anything to connect Parka with certain 

 rounded and ovate vesicles of larger size found detached in the same beds, and which 

 have been noticed by Mr. Powrie and Mr. Reid. These are not improbably vegetable, 

 and possibly sporocarps or iudusia, but I have not found them to show any structure. 



With reference to the second question, we cannot connect these bodies with Psilo- 

 phyton, whose fructification is well known, and the only other plants on the slabs are the 

 rugose stems above referred to and the narrow Zos^era-like leaves which would seem to 

 have constituted their foliage. These plants occur both in the Rescobie and Myreton 

 specimens on the same slabs with Parka. In carefully examining the slabs I find a num- 

 ber of masses of Parka placed at the extremities of branches or fragments of branches of 

 these rugose stems. This apposition may be accidental, but it occurs so frequently as to 

 give some probability that it indicates an organic connection. 



Putting the parts together in accordance with these facts, we may suppose Parka 

 decipiens to be the fruit of an aquatic plant having strong rugose but not woody stems or 

 rhozomes, producing numerous branches, those which were fertile, and perhaps nearer 

 the base, supporting clusters of Parka, those which were barren producing long grass-like 

 floating leaves like those of Zostera. The affinities of such a plant would be with modern 

 rhizocarps, though a peculiar and exaggerated form. 



In the meantime Parka decipiens may be accepted as an addition to the vastly profuse 

 rhizocarpean flora, which we know from American examples to have been present in the 

 waters of the Devonian period, as the author has shown in previous publications.' It 

 seems possible that the plant formerly described by the author as Cordaites auguslifolia,- from 

 the Brian of G-aspé may be allied to Parka, though only its leaves and stems are known. 

 Many rugose stems similar to the Scottish specimens have been noticed in both the Lower 

 and Upper Brian of Gaspé and the Baie des Chaleurs, and in both localities patches of 

 compressed vesicles larger than those of Parka, and in dense, closely packed masses, have 

 been found associated with these ; and at Gaspé I found, in 1868, a group of vesicles 

 similar to the Scottish specimens, but smaller. It is, therefore, probable that forms of 

 this kind existed on both sides of the Atlantic in the early Devonian. 



It is further to be observed that as we know the sporocarps of Prolosalcinia of Ohio 

 and of Brazil only as detached individuals, we cannot be certain that these may not ori- 

 ginally have been attached together in groups like Parka, and we do not yet know with 

 certainty the nature of their vegetative organs. In the meantime the facts above stated 

 should serve to guide investigation with respect to these interesting plants on both sides 

 of the Atlantic. 



It is proper to state that these new developments add to the evidence to which I have 

 referred in my papers on Protosalvinia,^ and in " The Geological History of Plants," ' in 

 favour of the great development of the rhizocarpean type in Palajozoic times. The enor- 

 mous quantities of sporocarps and macrospores in the Upper Brian shales of Ohio and 



' " Geological History of Plants," p. 48 el seq. ; ' Transactions Chicago Academy,' vol. i, No. 9, 1886. 



^ I do not now regard the Gaspé plant as of this species. 



"Canadian Record of Science,' 1883; 'Bui. Chicago Academy,' 1886, p. 105. 



* 1888, London and New York. 



