TUAXSACTIONS OF SECTION K, 809 



In one of the slides these sporangia were coherent in a cluster of six to eight, 

 and bore a very striking reseoiblance to the Calymmatotheca type, which has 

 been so far only known in the lorra of impressions. 



These sporangia are associated with petioles and other fragments of Lygino- 

 dendron oldMmiuvi, 



4. On some New Features in relation to Lyginodendron oldhamium. 



By J. LoMAx. 



Since the year 1872, when the late Dr. AV. C. "Williamson published his 

 Memoir, part iv., in the ' Trans. Roy. Soc' on Bictyoxylon, Lyginodendron and 

 Meterangiinn, this genus of fossil plants has undergone a good many alterations or 

 modifications at the hands of the palreobotanist. 



In the Memoir above quoted Lyginodendron oldhamium is described as an 

 unbifurcating stem, but giving off very small branches (see p. 387, figs. 14 and 16, 

 PI. XXV. loc. cit.), and in some respects similar to a Lepidodendroid stem. 



In his Memoir, part vi. 1874, he describes fully a stem to which he originally 

 gave the name ni Edraxylon, but which now appears under the name oi Rachiop- 

 feris aspera. (See part vi., figs. 1 to 1-3, ' Phil. Trans. Roy. Soc.,' 1874. Fig. 13 

 he de.scribes as a petiole with pinnules and leaflets.) 



In Memoir, part vii. 1875, he describes a series of specimens under the name 

 of Kaloxylon Hoolceri. (See figs. 23 to 38, PI. 5, G, and 7, loc. cit.) 



The small branches which Williamson described in 1872 under the name of 

 ' Kaloxylon ' have since been proved to be the roots of Lyginodendro)i. 



In the case of Jiac/riopteris aspera, it is now well known that it is the leaf- 

 stalk or petiole of Lyginodendron, whilst the fossils known as S'p/ienojjferts 

 Hiininghausi and Sphenojifcris disfans are the foliage of Lyginodendron, (See 

 Williamson's Memoir, 1890, part xvii. ; also Williamson's and Scott's Memoir, 1895, 

 part iii.) 



In most cases when this stem has been described it is as an unbifurcating or 

 unbranching stem. 



In Dr. D. H. Scott's ' Studies in Fossil Botany," p. 321, he says that ' No stem 

 has ever been observed to branch; it does not follow that it never did so, but 

 branching must have been very rare.' For some time back I have had various 

 specimens in my possession which I have collected from Duh-sgate, near Tod- 

 morden, Lancashire, which to me have had the appearance of giving ort'a branch; 

 but during the last few months I have discovered two specimens which have set 

 all doubt on this point at rest. Both specimens are in a iirst-class state of 

 preservation. 



The first of the two specimens was collected in March last, and is now in the 

 possession of Dr D. H. vScott; it shows a series of twenty-four sections cut 

 transversely. The series runs through two internodes ; the first shows the leaf- 

 trace just about to leave the periderm, and the stele is commencing to take a 

 crescent-like form on the opposite side to that of the leal-stalk which is being 

 given off, and so on till the stele of tlie branch and stem are completely separated, 

 and till the second leaf-stalk is given off. There are also several roots given oft', 

 one or two of which dichotomise immediately after leaving the cortex. 



In the above stem there are a series of four double leaf-trace bundles. The 

 entire diameter of the stem is 4' inch and about 3 inches long. 



The last specimen I have discovered is also from Dulesgate, from a mass of 

 small nodules, one part of which contained a quantity of foliage, broken pieces of 

 iSigillarian bark, Rachiopteris bihractensis, and various other fragments. 



In this there is a series of five leaf-trace bundles. This was about 1^ inch 

 in diameter and about 4 inches long. 



The stem is not quite whole ; a portion of one side is missing, but the develop- 

 ment of the branch is seen quite perfectly. 



The above two specimens show that Lyginodendron had a branching stem : also 

 that the branch was given off" in the one case between two leaf-stalks and iu close 



