30 
they are large and long. This is particularly the case in the Didy- 
mocarpi and Leptocarpi and has led to the founding of spurious 
species on this character. 
No fossil species of Astragalus have been certainly reported, 
while the geographical distribution points clearly to the Glacial period 
as the beginning of the primitive species. 
Doubtless there will always be discussion as to the indigenous or 
foreign origin of some species but generally speaking there is no 
reason to seek a foreign origm for any not known to be continental. 
Those species which are now contiguous in the Arctic, the Alpini, are 
of European or Asiatic origin, as well as the Hypoglottides, and prob- 
ably the Uliginosi. But the Hamosi of America though closely re- 
sembling those of Europe show no common origin with them but 
come down through the Flexuosi, or A. nitidus, and from the Hamosi . 
come the rest, such as the Leptocarpi, Micranthi and Didymocarpi, 
though the latter may have come from the Hypoglottides. If the 
line of descent were not clear then we might be justified in seeking 
a foreign origin, but though there are abnormalities and gaps yet as 
a whole the line of descent of all the species is well marked. While 
to connect our anomalies with species of remote regions in Europe or 
Asia would require far more unscientific guessing and stretching 
of ecological laws than the facts warrant, For example A. succum- 
bens has no near relative and yet its relation to the Hamosi ‘3 
evident as well as with the Galegiformes. The Uliginosi are evidently 
ef common origin with A. nitidus and presage the Preussii. The 
higher groups the Leptocarpi, Micranthi and Didymocarpi are the 
most differentiated and both ecologically and geologically have been 
subjected to the most changes of climate and the greatest migration, 
as shown by their present distribution and annual habit. We have 
hardly a suspicion of the spinose Astragali of Asia. A. jejunus 
and humillimus have rigid petioles as well as A. sesquiflorus, but 
never spinose. The adaptation of the species to climatic conditions 
is perfect as is that to soil whether the soil contains 1 per cent of 
soluble alkalies, or is acid as in the alpine meadows, or is barren 
clay as in the Navajo Basin, or the gravelly plains of Colorado and 
Montana, or the cooking-hot deserts of Arizona and Mexico. There 
is no evidence of present progressive differentiation, not even in 
such a cosmopolitan and variable species as A. lentiginosus, There 
is little hybridization and little evidence of fertility of hybrids. The 
species are easily definable though some are variable. The limits 
of a few species are uncertain, particularly Mexican, from lack of 
information. In fact many of the newer Mexican species are merely 
tentative, for there are at least a score still to be discovered which 
are liable to alter our conception of the limitations of the Strigulosi 
and Micranthi. 
LIFE ZONES. 
The prime factors in determining life zones are humidity (which 
includes that of the air and soil), temperature, soil, and light. Since 
the light is a matter of exposure, it varies locally, the same is true 
of the soil. Humidity is at the basis of all plant growth and should 
form the basis of all classification but unfortunately it, varies with 
altitude, exposure, drainage and many other things and so Tem- 
- perature is alone left to govern our plant zones. This varies with 
. altitude and latitude. We first find what are the actual plant zones 
