T. Holm — Erigenia oidbosa. 67 



developed during a season ; this is terminal, but it is not unusual 

 to find two or three more, which are lateral and developed 

 from the axils of the scale-like leaves. Whether the plant 

 produces flowers more than once is a question, which we have 

 not been able to answer, since no traces of flowering stems 

 from previous years were observed upon the large tubers, a 

 fact that does not, however, exclude the possibility of the plant 

 having bloomed before, inasmuch as the stems are so very 

 weak and evidently fade away entirely. The tubers of fruit- 

 ing specimens showed no signs of losing their vitality, but 

 were, on the contrary, in perfectly healthy condition and still 

 containing considerable deposits of starch : hence we suppose 

 that Erigenia may be considered as a polycarpic plant. 



While Erigenia is quite interesting at the seedling- and suc- 

 ceeding stages, when compared with the majority of the 

 Umbelliferw, hitherto examined from this viewpoint, there 

 are, furthermore, some peculiarities in its internal structure, 

 which may be mentioned in connection therewith. Moreover 

 it has not been possible from a mere superficial examination to 

 decide whether the tuberous body is a true tuber : a stem-part, 

 or simply a swollen root. The fact that we found no root- 

 hairs on the tuberous part, but only on the filiform part, does 

 not seem sufficient for separating them as morphologically dis- 

 tinct from each other, the former to represent a stem, for 

 instance the hypocotyl, the latter a root. An anatomical study 

 is necessary to decide this question, and since Erigenia does 

 not appear to have been examined anatomically, it may be well 

 to add some other details in regard to the position of its oil- 

 ducts for instance, which are very distinct and well developed 

 in this genus. 



If we examine one of the filiform roots of a tuber in its 

 third year, we find the following structure : Epidermis, this is 

 almost without roothairs, but many of the cells exhibit a slight 

 extension of the outer wall, though not enough to form what 

 is generally termed as " papillae " ; the cortical parenchyma is 

 thin- walled a^id composed of a few (four or five) strata without 

 deposits of starch, and, furthermore, neither lacunes or oil- 

 ducts were observed in this tissue. The endodermis (E in 

 fig. 5) is thin-walled and shows the dots very plainly ; inside 

 the endodermis is a pericambium of only one layer which sur- 

 rounds the leptome and hadrome. 



The hadrome constitutes a diametrical band of vessels, cor- 

 responding to two hadromatic rays of which the outermost 

 (the first developed) are narrower than the inner. On each 

 side of this single row of vessels is a group of leptome and 

 separated from these by a stratum of thin-walled conjunctive 

 tissue. Two, in transverse section, rhombic ducts (D) are 



