NOTES ON NEW ZEALAND FERNS. 365 



and Bicl-sonia Janata have creeping caudices, which anchor them- 

 selves to the ground with rootlets as they run. As both these 

 ferns grow high wp the mountains, so that they are buried in snow 

 all the winter, one has few oi3portunities of observing them. I 

 therefore got a number of plants of each last autumn, and planted 

 them in a cool shady part of my garden^ where they are all grow- 

 ing well. From watchmg then growth I am satisfied not only 

 that they can never have a true erect caudex, but that even Mr. 

 Henry and myself were deceived as to the soft portion of its 

 termination standing erect. The whole caudex to its very end is 

 prostrate, but the stipes of the crown rise at right angles (or 

 rather at obtuse angles, as the plants grow on deeply inclined 

 places) to it, and so closely together as to give the idea of the end 

 of the caudex being pointed vertically. The continued horizontal 

 growth of the caudex is, however, conclusively shown by the fact 

 that, each sjDring, the new crown of fronds rises from beneath the 

 side of the old one, and not from out of its centre, proving that 

 during the year the apex of the caudex has extended beyond the 

 cncle of the old stipes. The caudex, as I mentioned in a previous 

 letter, often creeps to a distance of thu-ty or forty feet from its 

 original root, and I can only account for the circumstance of its 

 habit in this respect remaining so long unnoted from its small 

 diameter and its being buried under the rotted fronds and other 

 decaying vegetation, and from the fact that the plants grow in 

 places which are seldom visited by persons likely to notice the 

 peculiarity, except very occasionally towards the end of summer or 

 early autumn, when the snow has melted from off the mountains 

 and the old crowns have withered and fallen. I think, however, 

 that I know how the mistake of supposing that these ferns have 

 occasionally a true erect caudex has arisen. When, in creeping 

 along, the caudex comes in contact with a fallen tree, it rises over 

 it, and descends to the ground again on the other side ; and in the 

 same way, on reaching the face of a precipitous bank, or face of 

 rock, it climbs it, and resumes its horizontal course on the top. In 

 either case, the caudex would be, for a time, actually erect, and a 

 person ignorant of the creeping habit might easily fail to notice 

 the fact that it was actually clinging by rootlets to the log or face ; 

 and the more so as, in this case, the natural growth of the crown, 

 at an angle to the caudex, would appear to have arisen from its 

 being forced out of the perpendicular by the obstruction, and would 

 to a great extent cause the fronds to hide the caudex. The caudex 

 of Lomaria jn-ocera constantly grows in the same manner, and on 

 steep faces often (I think more often than not) actually descends. 

 So usual indeed is the creeping habit, particularly in large plants, 

 that many people fancy this fern has a creeping rhizome. In the 

 description of Alsophila CoJensoi I notice a misprint in the * Sjmopsis ' 

 of inches for feet in respect of the length of the frond. Dicksonia 

 squarrosa grows fully twenty-five feet high, and the caudex, which 

 is very slender and hard, is covered by the bases of old stipes 

 broken off at a distance of from eight to twelve inches from it. No 

 weki grows on this fern. With us the stipes, rachis, &c., and the 



