CASTS OF CALAMARIEAE 47 



jection is almost constantly found on each ridge, 

 immediately below the node. 1 Various explanations 

 have been given of these projections, but a remarkable 

 specimen in the Williamson Collection has settled the 

 question. This specimen is a medullary cast of the 

 base of a branch. It owes its interest to the fact that 

 we not only find the little projecting bumps in their 

 usual position, but that many of them are represented 

 by regular spokes, radiating out for a long distance 

 from the cast. Williamson found, in the specimens 

 with internal structure preserved, that there was often 

 a radial canal passing through the parenchyma of each 

 principal medullary ray, near its upper end, which is 

 sometimes dilated. These " infranodal canals," as he 

 called them, extend through the entire thickness of the 

 secondary wood. In other cases, where there is not 

 a definite empty space, or canal, we still find that below 

 the node the rays are dilated, their tissue in this part 

 differing from that of the rest of the ray (see Fig. 8, 

 i.e., and Fig. 9, c). The dilated, infranodal portions of 

 the rays extend far out into the wood, even when the 

 rest of the ray is bridged over by intercalated tracheae. 

 Williamson's explanation was that these canals, 

 whether natural or, as is more probable, left by decay, 

 had, like the pith itself, become filled with mineral 

 matter, giving rise in ordinary cases to elevations on 

 the cast, which fitted into the hollow canals. In the 

 remarkable specimen referred to, the canals had been 

 filled throughout the whole thickness of the wood, and, 

 when the wood itself perished, the mineral casts of the 



1 Clearly shown in Figs. 2 and 3, pp. 17 and 18. In Fig. 3 these 

 prints are marked i.e. on the right. 



