98 DEVELOPMENT OF OSMUNDA REGALIS. 



to one lobe, the other to the left to the other lobe, when they each 

 agam divide into four secondary veins, and from these, others fork 

 off towards the margin. The two xDrincipal veins, however, only 

 subdivide each in its own lobe, and their respective forks do not 

 ever cross an imaginary line fi-om the notch to the footstalk. Tw^o or 

 three similar frondlets may grow from the same point of the pro- 

 thallus, and each will send down a rootlet or two, but this is all 

 the growth that is made the first year. There is no branching of 

 any kind, but only a single leaf-like expansion at the summit of 

 the single hair-like footstalk. 



The second year there appears from the same rootlets a rather 

 more developed frondlet, arising from a small winter bud, which 

 remains after the first growth has withered. This second year's 

 frondlet is more distinctly reniform, and occasionally show^s itself 

 in a form passing towards the next year's product. In other 

 respects it differs but little from the first year's, excepting in its 

 larger size and more developed lobes. 



The third year's frondlet is generally decidedly ternate, or 

 trilobate, but occasionally a frondlet is produced which is a kind of 

 passage form from that of the previous year, or sometimes indicating 

 the outlines of the next year's. In what we may call the normal, 

 third year's form, i. e., the trilobate frond, the princii^al vein 

 divides very shortly above the junction of the stem into three other 

 principal veins, which pass severally towards the apex of each 

 lobe, and these again subdivide in forks towards the apex and 

 edges, but, as in the previous instances, not passing over the 

 imaginary line dividing the lobes. Another vein is given off, also, 

 from each of the side veins, near their base, passing in a down- 

 ward direction to the base of the lobes. 



In the fourth year we have a much more complicated arrange- 

 ment, showing a decided and gradual approach towards the perfect 

 plant. One large tripartite frondlet is first thrown up on a long 

 slender hair-like stem, and afterwards other shorter fronds. The 

 latter bear at their summit a similar- shaped tripartite pinnule to 

 the first-formed one, only much, smaller, and below it a x^air of 

 pinnules very similar to the ordinary pinnules of the perfect plant, 

 one on either side of the rachis, and are more or less auricled at 

 the base. The rachis of the first-formed frond forms a mid-vein 

 which runs up through the centre of the terminal pinnule, and 

 throws off, close to its junction with the laminae, two other side 

 veins which run towards the apex of the other two pinnules, with 

 a short descending branch to each, as in the third year's form. 

 Sometimes the fourth year's plant is of a more luxuriant growth, 

 and exhibits a kind of transition form towards the next year's. 

 This often bears on its first-formed frond a ])e)fectli/ trifoliate set of 

 pinnules, the terminal one being irregularly trilobate. 



The fifth year's bears a similar i^rimary-formed frond, perfectly 

 trifoliate with a trilobate irregular terminal pinna, and lower down 

 on the rachis another very similar frondlet, having the terminal 

 pinna more regular. The second and subsequent fronds, which 

 are more numerous than in the fourth year, generally consist of a 



