1892.] of the Fossil Plants of the Coal-Measures. 471 



rows are defined as being more numerous, and the scars quincuncially 

 arranged. In Vlodendron each such scar is farther surrounded 

 by a large circular, or oval, and very characteristic disk. The author- 

 shows that the essential and homologous structure in all these fruit- 

 bearing branches is a small circular area, forming the summit 

 of a larger or smaller conical arrested branch which was covered 

 with leaves. This small apical area represents the part at which 

 the deciduous fructification was organically united with its sustain- 

 ing branch. Each such branch was supplied with a distinctive 

 form of vascular bundle, which differed alike from those larger ones 

 seen in ordinary vegetative branches, and from the smaller ones 

 passing outwards to the leaves ; this bundle is always abruptly 

 broken off at the extreme apex of the fructigerous tubercle in a way 

 demonstrating that it was formerly prolonged into some deciduous 

 appendage which is rarely preserved in situ. But Ulodendron has, in 

 addition, surrounding each of these fruit-bearing points, a flattened 

 surface, the size of which was mainly dependent upon the age to which 

 the tree had attained when it perished. This orbicular surface was 

 primarily covered with ordinary leaves, normally arranged, but the full 

 development of which was arrested by the pressure of some external 

 agent. The author concludes that the central fructigerous point was 

 homologous in all these cases, and that the variations seen in them 

 arose largely from the degree of prominence attained by the arrested 

 lateral branch. When that prominence was sufficient, the cone-like 

 fruit was pedunculate, and no disturbance of the surrounding leaves 

 was produced ; but when that elevation was small, or almost non- 

 existent, the cone was practically sessile, and, as it grew, its expanding 

 base crushed down the leaves which it covered and thus produced the 

 large flattened disk characteristic of Ulodendron. These two names, 

 Halonia and Ulodstidron, have no longer any generic value, but the 

 terms Halonial and Ulodendroid may be conveniently retained as 

 adjectives applicable to appropriate specific forms. 



The author applies these conclusions to his younger specimens of 

 L. Harcourtii, and shows that many of them were fructigerous in the 

 Halonial form. 



The organisation of the leaves of some of these Lepidodendroid 

 plants has been re-examined. As is well known, on the leaf -scars 

 alike of Lepidodendron and of Sigillaria, each scar left on the bark 

 after the fall of the deciduous leaf had three minute points impressed 

 upon its surface. Brongniart regarded each of the three as represent- 

 ing the entrance of a leaf-trace into the leaf, and very recently some 

 other observers have arrived at the same conclusion. The author, long 

 ago, showed that the central spot alone represented the vascular 

 leaf- trace, the two lateral ones being merely cellular structures, but 

 the details of which were very imperfectly known. These new speoi- 



VOL. L. 2 K 



