16? 



C. rosa marina, it elongates soon after the new biade is fairly well 

 formcd so that the adult internodes are two or more timcs longer 

 than the diameter. One, and a very important set of differences 

 between the three species, then, are these : in C. rosa marina, the 

 young internode is at first very short, but elongates with age: in 

 C. simplex, the young internodes are very short and hardly elongate 

 at ali with age; while in C. subulifera, the young internode is long, 

 even before Ihere is any trace of the new biade and it elongates 

 also with age. Or, in anothcr way, we may, look at the relative 

 lengths of the internodes. The distance between the adult laminae 

 is recorded on the stipe and its branches by the annular scars left 

 after the erosion of the laminae after maturity. The distances between 

 the scars mark the internodes. hi C. rosa marina and in C. subuli- 

 fera, the adult internodes are two to four times as long as the dia- 

 meter of the stipe, while in C. simplex, the adult internodes are al- 

 ways shorter than the diameter of the stipe. Postels and Ruprecht 

 say that the internodes are twice as long as thick in C. rosa marina 

 and four times as long as thick in C. Sitchensis, but the proportions 

 vary as much as that even in the same specimen. Consequently 

 this distinction as to exact proportions between the length and the 

 thickness of the internodes seems to be of little specific value, but 

 the contrast between the two species with long adult internodes and 

 that with short, as given above, is Constant and conspicuous. The 

 condition of the adult biade, as to whether it is split or entire, va- 

 ries so much according to locai conditions and vigor of the plant 

 that it cannot be depended upon either. In connection with the phe- 

 nomena descrihed above, stili another distinction may be pointed 

 out, viz.: that the in C. rosa marina and in C. simplex, the laminae 

 are peliate, becoming perfoliate only upon the appearance of the 

 new lamina, while in C. subulifera, the laminae nppear to be orbi- 

 cular and perfoliate from the very beginning. 



Besides the agreement existing between these three species as 

 to general morphology, there is also a dose agreement in the histo- 

 logical details. The structure of C. subulifera has been descrihed 

 somewhat in detail by Freeman (Minnesota Botanical Studies, sec. 

 ser., pt. II, pp. 175 et seq., pi. 17, 18, 1899) while Postels and Ru- 

 precht (loc. cit., pi. XL, ff. 85-88) give some fair details of a cross 



