139 
in so deep water, that it impossibly can have reached the 
surface with the pistils. We have gathered it fruiting even 
in meter-deep water or more. 
It also occurs macrocarp: var. obliqua (Schur) Asch 
&Graebn. (= R. transsilvanica Schur), and with narrower, 
seemingly more curved fruits (R. roste/lnta Koch), but pro- 
perly with thinner, less developed putamen, originating per- 
haps from the chemical composition of the soil (want of 
calcium?). 
To judge from the distribution in the Baltic, the north- 
ern limit of R. mar. might coincide with the year-isotherm 
of + 4° C. To the south of this line it extends over the 
whole world through the deserts of Africa as far as to Cape- 
town. To the north, again, of this boundary in America 
(and Asia?) and in the Rocky Mountains (Britton and 
Brown: Nebraska to Brit. Columbia) we might have the 
distribution-area of Ritppia occidentalis Wats. 
As to R. spiralis it varies very considerably. A medi- 
terranean type is var. drepanensis (Tineo, as sp.) K. Schum., 
with capillary thin uni-nerved leaves. Anatomy of stem and 
leaves like that of R. maritima. A northern type, charac- 
teristic to the Baltic, with narrow and thicker, likewise uni- 
nerved leaves we propose to name: var. subrigida, the leaf- 
anatomy of which deviates as the fig. shows. Besides we 
have observed a broad-leaved type with stouter stem. Leaves 
nearly 1 mm. in width, in the very margins endowed 
with small vascular bundles and, consequent^', 3-nerved. 
Stem also with two primitive bundles in the bark consisting 
of only one vessel surrounded with 6—7 mechanic cells, by 
this approaching the new species below, which has well 
developed cortical bundles. This might suitably be named 
var. latifolia. — A South American type presents a little 
more tapering leaf-points. 
A very beantiful Ruppia is the following from Tas- 
mania: 
