4 
T. G. HALLE, 
(Schwed. Südpolar-Exp. 
ever, involve any confusion. On the contrary, it is when the limits of such species 
as may definitely become established have to pass right through one of the too 
widely delimited older species, that the trouble in nomenclature really begins. 
In comparisons of fo.ssil floras the same principle will be seen to hold good — 
indeed, to afford the only safe basis for discussion. By employing too great caution 
in identifying seemingly related species of two distant contemporaneous floras, the 
difference between these floras is, it is true, unduly emphasized. It is probably a 
recognition of this obvious fact that causes many authors to use very wide genera 
and species in their comparisons of different floras. But in all questions of the geo- 
graphical relations of fossil floras, it is clearly the resemblance rather than the differ- 
ence which should serve as a basis for discussion; this, because an apparent difference 
between two fossil floras may b ■ caused by many other factors besides isolation in 
different phyto-geographical provinces. Any one who has collected fossil plants at 
a number of localities in one and the same district, knows that in two places not 
necessarily very far from each other the flora may be very different, because of 
some edaphic factors or some circumstance connected with the transport or preserva- 
tion of the plant-remains. Even a fairly great difference between two floras, whether 
near or very distant from each other, cannot, therefore, be regarded eo ipso as a 
proof that these floras belong to different phyto-geographical provinces. Thus, if by 
too great a reluctance to identify species of two distant floras, the difference is ex- 
aggerated, this need not commit one to any wrong phyto-geographical conclusions. 
It is only when a large material from several localities in one district is seen to 
differ considerably from a correspondingly large material from another district, that 
the difference becomes a fact of importance and permits of drawing any conclusions. 
But if a large material is examined, there is no danger that the points of resem- 
blance would not become evident too, however narrowly the species are delimited 
and however much caution is taken in the identiflcation. Only, if the species are 
narrowly delimited, the eventually established relation between the floras will be 
much more reliable than if the opposite principle is adopted. To prefer to keep 
the species as small as possible and to give new names to such forms from a distant 
region which do not quite agree with the most similar forms from other parts of 
the world, is therefore to follow the safer and more critical course in these 
matters. 
In order to avoid the institution of too many species and to mark the resem- 
blance between similar but not identical forms in different regions or from different 
geological horizons some authors have made use of varieties or of a ternary nomen- 
clature. Cladoplilebis denticulata var. australis (MORRIS) (Seward 1904 <?, p. 171) 
and Cladoplilebis Roesserti groenlandica Hartz (1896, p. 228) may be mentioned as 
examples. It is difficult to perceive, however, what is to be gained by adopting 
