MancJiester ][Iemoirs, Vol. liv. (1910), No. II. 9 



is well preserved in six successive sections. In two it has 

 definitely four bundles in a quadrangular stele. The next 

 section has a more or less triangular stele with six 

 bundles, while the remaining sections have triangular 

 steles which appear to have three bundles. One of these 

 latter shows six sporangiophores. The other two series 

 showing the change have only four sections each, and are 

 less well-preserved. Subsequently I found the same 

 change exhibited in a series of five sections cut from a 

 Halifax cone (Manchr. Museum Coll., O. 288 — 292). 

 Four of these have triangular steles, the fifth a quad- 

 rangular one. This series was formerly described by 

 Hick (Hick '93A), and this point was noted, but it seems 

 to have been entirely overlooked since that time. A 

 series of three sections in the Manchester Museum collec- 

 tion (R. 107, 109, 1 10) have each triangular steles, with six 

 bundles in two cases, and seven in the third. 



In six sets of serial sections which I have examined, 

 therefore, the stele changes from the triangular to the 

 quadrangular type in four, and shows indication of a 

 change in a fifth (the last-mentioned above). 



It is perhaps noteworthy that in the best series of 

 sections we have, between the part of the cone with four 

 bundles and the part with three, one section in which 

 there are distinctly six. Some indication of the same 

 thing is found in the series marked C. Hence it may be 

 possible that six bundles are produced in the readjust- 

 ment of the stele in this intermediate region. This would 

 explain the occurrence of a small number of sections in 

 which there certainly are six distinct bundles, and some 

 cases in which the bundles are irregular. I had hoped to 

 be able to determine this point by means of serial 

 sections made by the Sollas method, but have so far 

 been unable to obtain a block containing a suitable cone. 



