320 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1919. 
seems likely to have been used more generally in former times, when 
steel tools were not available. 
Advantage is taken of the fact that bush lands can be burned over, 
even without cutting, if the vegetation becomes sufficiently dry at 
the end of the hot season. The dead bushes and small trees that 
remain standing do not interfere with planting, nor with the growth 
of the crop. With the first rains that moisten the surface soil, the 
maize plants shoot up rapidly and are well grown before the sur- 
viving trees or bushes are able to put out leaves or new sjDrouts from 
the roots. As the woody vegetation is dormant at the end of the 
dry season, the roots are not likely to be killed and the new sprouts 
that are sent up during the rainy season are sufficient to shade the 
ground and exclude the dangerous grasses. 
The hot valleys and parched lowlands where the uncut bush gets 
dry enough to burn are not places that would be considered very 
desirable for human habitation, but to avoid the labor of bush- 
cutting would be a very important consideration with primitive peo- 
ple. Probably the same system would be applicable in Yucatan, and 
a passage in Norman's Rambles in Yucatan states that burning was 
"the only preparation that the soil received prior to sowing it." 
This related to the country between Merida and Campechy, which 
Norman visited in April, 1842. 
Though planting in uncut bush is even simpler and easier than 
the regular milpa agriculture, it may not be more primitive, since 
the method would not be applicable to orginal forests, but only to 
secondary growth. It may be significant in this connection that the 
southern cities of the Mayas in Honduras and Guatemala, in the 
regions of heavier forests, were older than the cities of Yucatan. 
ARTIFICIAL GRASSLANDS AND DESERTS. 
That the agricultural operations of primitive man may change com- 
pletely the character of the wild vegetation and turn a dense tropical 
forest into an open grassland or a desert is a fact not yet appreciated 
adequately by students either of plant life or of human progress 
toward civilization. The biological considerations indicate that in 
its primal, prehuman condition the tropical and subtropical world 
had a general forest covering, and that tropical grasslands are essen- 
tially artificial. 
Grasses do not exist naturally in lowland tropical forests, being 
intolerant of shade and unable to compete successfully with the woody 
vegetation. Apart from special local conditions of salty soil, peri- 
odic floods, or fires that in some regions are kindled rather frequently 
by lightning, there is nothing to keep the woody vegetation from in- 
vading and becoming established in any region where the rainfall is 
