MancJicstcr JMcmoirs, Vol. Hi. (1908), No. 4. 3 



secondary growth. If this be true these small Bothro- 

 deiidron cones cannot have produced the great Uloden- 

 droid scars of /A punctatuiii. 



The largest definite strobilus of a Lepidodendroid 

 tree with which I am acquainted is Lepidostrobiis Broivnii, 

 the internal structure of which is preserved. The diameter 

 of this cone is two inches, and Ulodendroid scars are 

 known about four times this size. 



The fact that certain species which bear Ulodendroid 

 scars are known to have cones attached to the ultimate 

 branches suggests that the Ulodendroid scars could not 

 have directly borne cones, as it is exceedingly unlikely 

 that a Lycopod would carry its cones in two distinct ways. 

 There is thus a certain amount of evidence that " Ulodcii- 

 droiT could not have had sessile cones attached to its scars. 



The principal specimen on which I depend for 

 support to the branch theory is contained in the Man- 

 chester Museum. It is represented in Plates I. and //. 



This specimen is a cast in very fine Coal Measure 

 sandstone of one Ulodendroid scar probably referable, on 

 account of its ovoid form and eccentric umbilicus, to 

 Bothrodendroii pnnctatinn L. & H. The specimen is not 

 localized and not registered, but Professor W. Boyd 

 Dawkins informs me that it is probably from the Middle 

 Coal Measures of Peel Uelph Quarry, near Bolton. It is 

 to be particularly noticed that this scar apparently formed 

 the centre of a concretionary area in the sandstone, and 

 that in parts a certain amount of structure is preserved in 

 iron carbonate. 



One of the remarkable characters of this specimen is 

 that in the obverse half (see Figs. 3, 4, Plate II.) the 

 umbilicus is represented by a cylindrical hole 18 mm. 

 deep and 8 mm. in diameter. 



This hole is surrounded by an upstanding ring of iron 



