from the Middle Coal Measures, near Rochdale. 75 



represent the points of rupture of the vascular strands, 

 rather than the insertion of the bracts, and, if so, the 

 variable position would not be difficult to explain. 



The appendages to the axis consisted originally of 

 whorls of bracts arising at the nodes, but whether or not 

 they were connate at the base cannot be determined 

 with certainty. The presence of a midrib is also 

 uncertain. The specimen shows that the bracts stood 

 ■out from the axis horizontally for about 3 or 4 mm., 

 and then turned upward and outward at an angle 

 •of about 120 , in which direction they continued for 

 upwards of 10 mm. (2-5ths inch). The erect portion 

 of the bracts appears to have been very narrowly 

 lanceolate in shape, and to have had a somewhat 

 elongated apex. The number in each whorl was not a 

 large one, probably not more than 20; but an exact 

 •determination is impossible. 



The sporangia associated with the bracts are not 

 well preserved, but the appearances point to the con- 

 clusion that they were ellipsoidal in form, with the long 

 axis of each slightly inclined to the vertical and towards 

 the free portion of the bracts. The sporangial walls 

 exhibit something like longitudinal foldings ; but whether 

 these existed in the living state or have arisen during 

 fossilisation it is impossible to say. 



On the most critical point of all, viz., the relation 

 of the sporangia to the bracts and the axis of the 

 spike, the evidence of the fossil is practically nil. In 

 no part of it has it been possible to make out the 

 presence of a sporangiophore, a structure whose charac- 

 teristics are of the first importance as a guide to natural 

 affinities. On the other hand, there is nothing in the 

 fossil to justify the definite conclusion that such a 

 structure did not exist. All that can be said is that 

 the sporangia seem to be sessile on the whorl of bracts 



