Lee 
With what may the Trigonella-formation be most closely 
compared? In many places its physiognomy resembles a 
semi-desert, an Artemisia-desert, for instance, or one formed 
by suffrutescent Salsola species, (see PAULSEN, page 69); — and 
SCHIMPER, Who, however, only knew the vegetation of Pamir 
from photographs (fig. 449—455 in his book), says, (page 792), 
that the flat valleys have the character of deserts, just as 
does FEDTSCHENKO, (see above, page 28). Compared with the 
Transcaspian deserts these differ physiognomically by con- 
taining the many large cushions of the two Acantholimon- 
species. Cushion-plants (Anabasis aretioides) are, though, to be 
found in the Sahara. Perhaps another difference is to be 
found in the fact that hemicryptophytes, and not chamae- 
phytes, play the leading part, in any case in quantity of 
species, that thus woody shoots are less conspicuous here than 
in deserts. However this difference is not vital, as Transcaspia 
has a considerable quantity of xerophytic hemicryptophytes, 
even though their importance is less than in Pamir. 
A third difference is that there are no vernal-flowering 
species here, while in low deserts the majority of the species 
flower in the spring; the lateness and coldness of the spring 
is the reason for this. 
A fourth, and in my opinion conclusive difference, is 
the fact that the Trigonella-formation is poor in therophytes, 
annual species, while low-lying deserts are characterized just 
by the large numbers of therophytes found there. Salt- 
deserts are the single exception; this is due to a special 
edaphic cause, the soil s content of salt, while in Pamir the 
cause is doubtless climatic. WARMING, too, (1909, page 251), 
emphasizes this, and, quoting various authors, shows that the 
percentage of annuals diminishes with the increased altitude, 
(also with the geographical latitude), and remarks that the 
cause of this lies in the shortness of the vegetation-period, 
and the low temperature, only permitting a few annuals to 
complete their development and set ripe seed. 
The small number of therophytes present is the main 
reason why the Trigonella-formation may not be considered 
a desert-formation. When f. inst. SEMENOW says the vegetation 
of Pamir has steppe-character, and WARMING (1909, page 260) 
