o/" Ulodendron to Lepidodendron, &tc. 241 



That the appendicular organs were caducous cones seems 

 most probable ; but I have not seen evidence sufficiently clear 

 to decide positively whether they were sessile or stalked. From 

 the evidence before us, however, and taking into account the 

 morphological significance of the attached organ, I have a 

 strong bias in favour of the opinion that the appendicular 

 organs were sessile cones. The view advocated by Stur that 

 they were bulbils does not appear to me to be at all probable, 

 and against it Schimper has stated sufficient objections*. 



Before meeting with the specimens which form the subjects 

 of figures 2 and 9, it had often been a mystery to me why we 

 never found Ulodendroid scars on small stems, especially as 

 Halonian branches of Le^ndophloios scoticus, Kidston, about 

 half an inch in diameter, are frequent. This difficulty is 

 quite cleared up by an examination of the specimens nos. 8 

 and 18, of which the portion drawn in PI. IV. fig. 2, is 

 from a stem 4| inches broad (specimen No. 3) , and that in 

 PI. V. fig. 9 (specimen No. 18) from another over 3 inches 

 wide. It appears, then, that in the so-called Ulodeiidi-a it was 

 only the older stems that bore lateral cones. 



It has already been pointed out that plants belonging to the 

 genus Rhytidodendron, Boulay, also possessed two opposite 

 rows of Ulodendroid scars f. It is true that Ulodendroid 

 specimens comparatively seldom show the leaf-scars well 

 preserved, and that on some of the described species of Ulo- 

 dendron they have not been observed ; still that does not alter 

 the fact that when well-preserved examples are examined 

 they show leaf-scars which conform to one or other of the 

 three genera Lepidodendron, 8igiUar{a, or Rhytidodendron^ 

 as already mentioned. But if the form of the leaf-scar is 

 of generic value in Lepidodendroiiy SigiHaria^ and Rliytido- 

 dendron, on what grounds can we ignore the value of the same 

 character in Ulodendron ? If, then, those plants with Ulo- 

 dendroid scars are to be excluded from Lepidodejidron^ Sigil- 

 larittj and Rhytidodendron, it will be necessary to form three 

 new genera for these plants — one for the Lepidodendroid Ulo- 

 dendra^ another for the Sigillarian Ulodendra, and a third for 

 the Bhytidodendroid JJlode.ndra. This view, however, I am 

 not prepared to adopt, ,as I think the plants find a suitable 

 and natural place in the genera Lepidodendron^ Sigillaria^ 

 and Rhytidodendron respectively. 



Branching of Ulodendroid Stems. 

 The Sigillarian as well as the Lepidodendroid species of 

 * See ante, p. 135. f See ante, p. 138. 



