MALARIA EXPEDITION TO THE GAMBIA 
27 
from the conditions which obtain at the end of the wet season, when I personally 
examined the swamp, I surmised that puddles are formed in the natural hollows of 
the streets all along the margins of the swamp, namely, Albion Place, Perseverance 
Street, etc. ; indeed I found, on arriving at Bathurst, remains ot puddles in these 
streets ; some were dried up, others still contained a little water, in which Anopheles 
larvae were present. Box Bar swamp presented some interesting features during 
the months of October and November. A long, partially constructed central channel 
runs the length of the swamp, opening by means of double sluice gates into the Oyster 
Creek at the back of the town. This channel contains a large quantity of water, which 
overflows on to the swamp after heavy rains. On either side of this channel the land 
gradually rises one to three feet to Portuguese Town on the one side and New 
Town on the other ; it is covered with grass and a few low bushes. This land has a 
very uneven surface ; there are very many hollows and depressions occurring at 
intervals amongst the grass, and also not a few large round holes three to four yards 
across, which may have been formed by the natives in obtaining sand ; also, on the 
borders of the swamp, especially around New Town, trenches occur choked with 
grass. Lastly, innumerable crab holes are present everywhere, and on to the surface 
old tins and calabashes are thrown. In the central mass of water I was never able 
to obtain mosquito larvae, undoubtedly owing to the enormous quantities of small 
fish which are always present in the water. But in October and November the 
hollows and depressions mentioned above as occurring amongst the grass contained 
water, which was cut off from the main channel, and thus unable to drain away, nor 
could small fish gain entrance ; it was thus to be expected that these pools should 
contain innumerable mosquito larvae. Three large pools in close relation to this 
swamp, situated together near the cemetery, must be specially mentioned ; these pools 
measured twenty to twenty-five feet across, and contained water in December, when the 
swamp pools proper had completely dried up ; they were used then for watering cattle. 
As Box Bar swamp dried up these pools became more and more infested with mosquito 
larvae, though one of them remained free for some time. In it the water was com- 
paratively sweeter than the other two, and small fish were present. As the dry season 
advanced the water in these pools became exceedingly foul, and enormous numbers of 
larvae, chiefly Anopheles, were found. A long, narrow pool occurs in this neighbour- 
hood, along the side of the cemetery ; this pool communicated with the creek, and 
is flushed with tidal water ; though the water present in it was foul, and had a dis- 
tinctly unpleasant smell, no mosquitoes were found breeding in it. 
Half Die. This swampy district is drained in a similar manner to Box Bar 
by means of a central channel opening into the creek by sluice gates. No mosquitoes 
were ever found breeding in the central mass of water. The swampy surface is much 
more level than Box Bar, and very little grass is present. Native huts are built close 
up to the margin of this swamp, practically on all sides, and it is in their compounds 
