difference between the estimates of the fir and kail areas 
by ground survey and photo-interpretation (scale 1:40.000) 
is very small (probably due to simple structure and compo- 
sition of vegetation) whereas in the case of mixed forests 
of Bastar the differences are significant. 
  
(iv) Difference in the mean height of 294 plots on the ground 
and the photo is negligible. However, the high magnitude 
of the standard deviation of the difference may be due to 
random differences in location of trees on the ground and 
photo plots and also due to the differences among the 
interpreters. 
(v) Lastly it can be said that black and white panchromatic 
aerial photographs can be effectively used in preparation 
of forest vegetation type and landuse maps, in estimation 
of area under different vegetation types, reducing the 
amount of ground sampling by forest stratification prior 
to undertake field work and to improve the total volume 
estimates. 
2. Landsat imagery in forest mapping and surveys: 
In order to keep pace with ever increasing requirement of raw 
materials for agro forest industries, a comprehensive, rapid, 
real time and repetitive assessment of the forest resources 
potential becomes inevitable. The use of frequent aerial photo- 
graphy for the assessment of large tracts of land is prohibitively 
costly. The recent remote sensing techniques, particularly multi- 
band sensing methods which are still in early stages of research 
with reference to forestry is expected to provide some solution 
to the problem. 
2.1. Material and Method of Study: 
With a view to find out the usefulness of Landsat imagery in 
forestry, the Landsat imagery (E-1219-0437-502 exposed on 26.2.73 
and E-1219-0437-502 exposed on 27.2.73) in bands 4, 5, 6 and 7 
on 1:1000.000 and in band 5 on 1:250.000 scale with reference 
to Mahadevpur Reserved forests of Karimnagar East Forest Division 
and to the Gudem reserved forests of Kagharznagar Forest Division 
respectively of Andhra Pradesh were studied and interpreted 
visually. The false colour print of the imagery pertaining to 
the second area of study on 1:250.000 scale was also studied. 
All the imagery were of good quality and free from cloud shadows. 
The general tone of band seven imagery was the darkest whereas 
that o band six was the lightest. Band five imagery was the next 
darkest. The tonal contrast among different vegetation types 
was observed and found to be best in band five imagery and poor 
in band six and seven imagery. Band four showed better tonal 
contrast than bands six and seven. And the false colour print 
imagery pertaining to Gudem reserved forests showed even better 
tonal contrast than band five imagery. As false colour and band 
five imagery offered best interpretation possibilities, enlarge- 
ments of band five and that of false colour were got made on 
1:250.000 for further study.