160 H. G. SIMMONS. [sec. arct. exp. fram 



other parts. The original material from Spitsbergen, on which Anders- 

 son established his Gatabrosa vilfoidea had only two-flowered spikelets 

 — hence the referring to Gatabrosa — but in the Copenhagen herbarium 

 I have seen as many as six flowers in a spikelet, indeed in rather large 

 specimens, that it would, perhaps, be better to refer to var. arenaria. 

 Very often the reduction goes still further and no panicles at all are 

 developed; this is rather often to be seen even in more southern lo- 

 calities, viz., in plants referred to var. arenaria, but in arctic localities 

 it becomes still more common, and generally only single inflorescences 

 are to be found. I cannot see that a plant thus continually connected 

 with another can rightly be looked upon as a separate species. 



Even if a single panicle is developed, that does not imply that ripe 

 seed will be developed, probably that is rarely the case, and the plant 

 is naturally cast upon vegetative propagation for its continued existence 

 and spreading. Consequently the creeping radicant stolons are develop- 

 ed to a considerably larger extent than in a form from more favorable 

 localities. Fragments of the stolons .will easily be spread by the ice, 

 which often reaches up to the growing-places of the plant and which 

 can detach parts of it and transport them. For the very few shore plants 

 of the Arctic Regions, this mode of transport may probably play an im- 

 portant part, as' they are able to sustain a longer immersion in salt water. 



My Ellesmereland specimens are all of them typical var. reptans 

 with small leaves apd densely radicant stolons, but from Fouike Fjord 

 I have specimens that come nearer to var. arenaria as they are con- 

 siderably larger in growth and have the radicant stolons less developed. 

 But the flowering was very scarce. In Ellesmereland the plant was 

 always sterile. 



It grew always near the shore-line, sometimes in gravelly places, 

 generally in clay or mud soil, around lagoons, &c. 



Occurrence. Hayes Sound region: Skraling Island (1386), Twin 

 Glacier Valley, Cape Viele, Cape Rutherford (450, 1155), Cocked Hat 

 Island (1287). South coast: Fram Fjord (1634); Harbour Fjord at the 

 "green patch"; Muskox Fjord ; Goose Fjord at 3rd quarters. Yellow Hill 

 and 4th quarters; Walrus Fjord. 



Distribution: East and West Greenland, Pribilof Islands, St. 

 Lawrence Island, Arctic Siberia, Novaja Semlja, Spitsbergen, Finmark. 

 The main species is spread on the shores of America, Asia and Europe, 

 and also found in salt-localities inland, however I doubt whether some 

 of the localities mentioned for Atropis maritima in Ledebour, F1. Ross., 

 4, p. 390, can really be referred to this. 



