Part II] Hole : Regeneration of Sal (SJiorea robusta) Forests 39 
Plate VIII, Fig. 1, shows the 1^-years-old seedlings in Plot XVIII, 
1 (c), on 31st January, 1919, and Fig. 2 of the same Plate shows the JJ- 
years-old seedlings in Plot XVII, II (c), on the same date. 
41. The chief points of interest are :— 
(1) The results in this experiment have been obtained on a fairly 
large scale, under forest conditions, by sowing broadcast and 
with no fencing of any kind. 
(2) The sowings on dead leaves throughout, both under shade 
and in the open, have proved a failure, the survivals in 
all cases being less than 2 per cent. 
(3) In every plot, both in the shade and in the open, the results 
on the dug soil have been distinctly better than on the undug 
soil, both as regards the number of survivals and the height- 
growth. In some cases, where the ground is fairly level 
and the soil surface has not been hardened, quite good results 
may be seen on the undug soil. When the ground, however, 
is not level, the seed is quickly washed off by the rain, whereas, 
in dug soil, the rainfall percolates in situ , the seeds are not 
washed away, a well-distributed and even stocking results 
and an equable distribution of moisture in the soil is secured. 
(4) In the area with the heaviest shade, viz. Plot XXV, the sowings 
have been practically a failure with less than 3 per cent, 
survivals on the dug soil. This indicates that clearing the 
undergrowth without interrupting the overhead cover is of 
little use, 
(5) In Plot XVII where the overhead cover has been completely 
removed on a strip 60 feet wide, the results are undoubtedly 
much the best with 25 per cent, survivors and a height-growth 
of 12 inches after 1J years. 
In considering this result, also, it must be remembered that, when 
this experiment started, the strip had been cleared for 2 
years, and a good deal of weed growth had already estab¬ 
lished itself. 
(6) The rather poor results in Plot XVIII, the cleared strip 100 
feet wide, are due to the bad growth in sub-plot lie. In 
this area a dense matted growth of grass (Eragrostis cyno- 
suroides and Imyemta arundinacea) had established itself 
which impeded germination and interfered with seedling 
development. This was due to the fact that the line had 
been cleared and exposed to full light for 2 years before the 
experiment started, it having been cleared first in 1915. 
With ordinary care, full regeneration can be secured in the 
[ 201 ] 
