348 



MOSSES AND FERNS 



CHAP. 



the growth of the prothalhum is exactly opposite to that of 

 the first rhizoid (bi-polar germination), and Kny ((5), p. 12) 

 lays a good deal of stress upon this, as distinguishing Osmunda 

 from the Polypodiaceae ; but it is not at all uncommon for 0. 

 Claytoniana, especially, to have the axis of growth of the rhi- 

 zoid almost or quite at right angles to that of the prothalhum, 

 exactly as in the Polypodiaceae. Where the germination is 

 truly bi-polar the exospore is pushed up with the growing pro- 

 thallium, and appears like a cap at its apex, but if the rhizoid is 

 lateral, the exospore remains at the base. 



In O. Claytoniana there are usually several transverse walls 

 A. _ 



Fig. 192. — Osmunda cinnamomea. A, Young prothallia; B, an older prothalHum, X260. 



formed before any longitudinal ones, but in O. cinnamomea 

 and O. regalis it is quite common to have the first transverse 

 wall followed by a longitudinal wall in each cell, so that the 

 four primary cells are arranged quadrant-wise (Fig. 192, A, 

 c). Rarely the first wall in the prothallial cell is longitudinal, 

 as is often the case in Equisetum, and sometimes the first divi- 

 sions are in three planes, so that a cell mass is formed at once, 

 as so often occurs in the Marattiace^e. Where a filamentous 

 protonema is formed, a two-sided apical cell is soon established 

 in exactly the same way as in Onoclea. Where the four quad- 

 rant cells are formed, one of the terminal ones becomes at once 

 the apical cell. 



