376 MR K. C. DAVIE ON 



elaborate precautions taken by leaf-traces to prevent undue contractions of their 

 general water-supply due to the departures of successive pinnae. 



The two factors which have more than any others determined the development 

 of the complications and simplifications of the margins of the Fern leaf-trace have 

 been the need of carrying up past the earlier pinna-trace-departures a water-supply 

 sufficient to meet the demands of the succeeding pinnae and the terminal portion of 

 the leaf. The need for supplying the separate pinna-traces has been met by the 

 giving-off of the parts of the leaf-trace nearest to those pinnae. As the leaves became 

 larger and more heavily pinnate, provision was made for the water-supply of their 

 increasing number by the development of the incurved hooks on the adaxial face of 

 the leaf-trace. The departure of the pinna-traces took place still from the most 

 conveniently situated part of the leaf-trace, but, as this was the back of a hook, the 

 supply was extramarginal in origin. Then there seems to have been a condensation 

 of the leaf- trace (perhaps in relation to a reduction in the size of the leaf), shown in 

 the appearance of " broken " leaf-traces, either of the binary or of the many-stranded 

 type. For some time the hooks were retained on the adaxial strands, but they 

 gradually disappeared, and the pinna-supply was given off simply from the margins 

 of the adaxial strands. Here and there arose again the need for carrying forward 

 water past the pinna-trace departures. The marginal type of supply had become 

 useful in its possibilities of extensive supply to the pinnae. And the difficulty of 

 carrying forward the water-supply to the ultimate parts of the leaf was solved by 

 pressing the abaxial strand or the abaxial portions of the adaxial strands into the 

 service. Sometimes the adaxial strands were reinforced from the median strand as 

 necessity arose ; at other times they were provided throughout their length with 

 incurved abaxial hooks. If one may risk another generalisation from these details, 

 it would seem that the Fern leaf in the course of its phylogeny had developed first 

 in respect of length, at the same time as its appendages increased in size. Then there 

 came a reduction both in the length of the leaf and in the size of the appendages, 

 the reduction in length preceding the reduction in the size of the appendages. By 

 that time the Fern leaf-trace had become thoroughly adapted to the needs of the 

 leaf, and the later reductions or amplifications in special cases made but little change 

 on the configuration of its adaxial portion, but mainly affected the abaxial part. 

 Indeed, in the marginal type of pinna-supply which occurs in the more advanced 

 Ferns we have the most improved type, and one which has proved to be the most 

 adaptable. 



In conclusion, I desire to express my thanks to Professor I. Bayley Balfour, 

 F.R.S., for granting me the privilege of obtaining most of the material 1 have 

 examined from the Royal Botanic Garden, Edinburgh, and for communicating this 

 paper; to Professor F. 0. Bower, F.R.S., and Professor D. T. Gwynne-Vaughan, 

 F.L.S., for much helpful criticism ; and to Dr W. T. Gordon, F.R.S.E., for assistance 

 in the preparation of the microphotographs. 



