304 



Copenhagen the vertical shoots liail an even denser whorl of 

 roots, wliich however were not very h)ng, bnt greatly hranched; 

 the lowest internode of all was here quite covered by mud. In 

 deeper water in the pond of the hardens the corresponding 

 shoots had but few roots. 



As a rule the roots are iinbraiichcd ; tliey are triarch or 

 diarch (fig. 2). The vessels, which are naturally fewer than in 

 the land-form (Schenck 1886 b, Taf. X, fig. 69), gave no wood- 

 reaction in those examined, neither with phloroglucin-hydro- 

 chloric-acid nor with potassium pernianganate-hydrochloric- 



ammonia. On Danish spe- 

 cimens examined the red 

 colouration with phloro- 

 glucin was extremely faint, 

 not more than observable. 

 It will very probably be 

 more apparent in some 

 specimens. Root- hairs 

 are wanting (Schenck 1886 b 

 p. 58), as in many water 

 and amphibious plants. 



Transverse sections 

 of the stalk differ from 

 the structure described by Sanio (1865 p. 184) and from that 

 of most of the Danish plants examined, by having fewer — 3 

 or 2 — circles of large air-canals in contrast to the 5 of Sanio, 

 but they are wider here (fig. 3); the condition is the same 

 whether the section is made across the lower part of the stalk, 

 which ^vould be submerged, or 2 — 3 cm from the tip, in the 

 flowering, probably emerged part (a very strong specimen from 

 Igaliko 60° — 61° N. L.). The most divergent are some small, 

 sterile plants, which I had the opportunity to collect, on an 

 excursion of the Swedish Botanical Society late in July 1909, 

 in a lagoon at Torne Trask 68°23' N. L., 342-1 m above the 



Fig. 2. Hippuris vulgaris L. 

 Root, transverse section. 



