MancJiCster Memoirs^ Vol. Hi. (1908), No. 4. 9 



into such a deep hollow, could be bent over laterally 

 without breaking oiT, I do not understand. I think, there- 

 fore, that one of the two following explanations must be 

 true. 



Either the organ is not really attached, or, it is 

 attached, but the Ulodendroid scar was in life almost flat, 

 and hence could not have been produced by pressure of 

 the deeply conical base of the organ. With regard to the 

 determination of the organ as a cone, I can only say, that 

 judging from the figure given b\' Thompson, it is 

 impossible to decide definitely as to its true nature. 



2. The second specimen consists of a single Uloden- 

 droid scar preserved on a sandstone cast, into which fits 

 the base of an appendicular organ. 



This organ is broken off very short, about v^ cm. 

 still remaining. 



It shews on its surface some ill-preserved scars, which 

 may be the worn bracts of a cone, or equally well ill- 

 preserved leaf-scars. A comparison of Mr. Kidston's 

 figures of the undeniable leaf-scars on the stem, and the 

 markings on the appendicular organ shows that the 

 differences between them are very slight, in fact, 

 Mr. Kidston admits that they are so. 



The character presented by this specimen, on which 

 most stress has been laid as shewing that the appendicular 

 organ, whatever its nature, was only attached to the 

 umbilicus, is that, with the exception of the central point, 

 the whole area of the scar is covered with a layer of coal. 

 It seems to me that this specimen can be completely 

 explained on the branch theory, if we suppose that there 

 was a branch shedding mechanism in connection with 

 which a layer of periderm was formed across the base of 

 the branch, cutting across the soft tissues of the middle 

 cortex into the wood ; such a la\'er of periderm would not 



