June 1, 1901.] THE TROPICAL AGRICOLTURIST. 
825 
cart hire, which ahows a great saving in using the 
tram, more >^sp8oiall.y as the tram track cost much 
lesti for upkeep than ihe whole road. 
I believe the 25 miles of rail laid in North Travan- 
core is the only instance of Ewing's one-railed tram 
being tried on a practical scale. 
Ml- Ewing's Agents are Messrs. Parry & Co. of 
Madras, from whom all informatioa regarding it can 
be obtaiued. 
It need hardly be mentioned that the writer has no 
interest whatever in it, further than having been the 
pm'chaser of the above installation. Briefly, the sys- 
tem consists of 
(1) A single rail attached to sleepers in the ordi- 
nary way ; 
(1) Trucks, with two or more double-flanged wheels 
running underneath the centre of them, tandem-wiae, 
so that, theoreticalli/, if a truck were properly loaded on 
level ground, it should balance itself on the railwheels 
just referred to ; and 
(3) An outrigger wheel of large diameter (four feet 
or more) made of very light materials and wide tyre 
(four inches or more) revolving on a hinged axle, 
fixed to the centre of the truck and projecting over 
one side, while a*'tached to the axle at the edge of 
the truck are two spiral springs, one above and one 
below it, that allow the axle to move up and down 
freely and compensate for any inequalities of the 
road. 
The object of the outrigger wheel is to prevent the 
truck from upsetting, and .it is not intended to carry 
any load, the whole weight of the truck and the goods 
on it should be carried on the rail wheels. 
SLEEPERS AND RAILS. 
The rails we laid down were 1-i lb to the yard, and 
they were dog-spiked on the sleepers 2 feet 3 inches 
and 4 feet 6 inches long by 8 inches wide and i inches 
thick, and we found t best to place alternately a 2- 
feet o inch slt-eper across the rail, and then a 4-feet- 
6-inoh sleeper, longitudinally, leaving a space of about 
15 inches between them. Ijighter sleepers would not, 
in the writer's opinion, do. 
These rails cost 111,700 per mile, including fish- 
plates and spikes C.I.F. at Madras ; and sleepers cost 
B250 to R600 a mile, according to the cost of sawing 
and distance of transport, while bending and rail- 
laying and rolling-stock should cost another R700 a 
mile, biinging up the cost of the complete tramway 
to about E3,000 a mile. 
Our trucks were 12 feet long by 6 feet wide, and had 
simple platforms running on two wht-els for carrying 
rice and tea chests, costing about B400 each, and 
built to carry a load of 1 ton ; but, of course, rails of 
heavier section can bo l;id to carry heavier goods, 
and the tvuck^ can be built of any ordinary form 
and have more th n twi wheels if necessary, in 
which case I think the extra wheels or wheel should 
have plain treads and no flanges as the friction on 
these double-flanged wheels is a serious drawback 
to the tramway. 
EXACT POWER REQUIRED TO HAUL, A TON. 
I intended to test, with an electric mo l^or-oar we had 
built, the exact power required to haul a ton of goods 
on the tram, compared with the power ri quired to move 
it in country carts ; and 1 hope this experiment will be 
carried out yet, as it would be valaable. It cjahl 
either be done with the electric motor or with cattle 
and an ordinaiy hauling bar with a weight-tester 
attached, and I believe that the efficiency of the tram 
will be comparatively lessened lohenever curves of small 
radius have to be nef/ocinted, because double flanges on 
the wheels involve much more friction than single 
fliinges, as the wheels are fixed in a straight line, three 
or four feet apart, and any deviarion of the rail trom a 
straight line involves considerab e friction on the rail at 
the four points where the flanges of the wheels bear on 
it. In ordin.iry railwav-wheels this friction is overcome 
by the shape of the tread of the wheels, and, if any means 
can be discovered for getting over this difficulty in 
Ewing's tram-wheels, the power required for haulage 
will be greatly lessened, and the life of both rails and 
wheels greatly lengthened. 
Mr. Kemlo designed two arrangements, one for 
swivelling one wheel to meet the curves, and ihe other 
for using four single-flanged wheels instead of two 
double-flanged wheels, but those have never got beyond 
the stage of drawings, so far, although they are well 
worth experimenting with. 
CURVES. 
We were led to believe that curves of 20 or 30 feet 
radius were suitable for our tram, but after seven miles 
had been laid it was evident that these sharp curves 
caused immense wear and tear to wheels, rails and 
cattle, so the writer reluctantly had seven miles of rails 
pulled up and relaid at curves of 60 feet radius as a 
minimum, and the grades of the road were at the same 
time corrected, the steepest part being one in 28, and 
I would not advise anyone to lay in a Ewing's tram at 
steeper grades than this, unless for a short portion 
of the total distance. In makii>g anew track for a tram 
of this description, therefore, it should be borne in 
mind that curves should be 50 to 60 feet in radius, and 
where a tram road has to be cut in a steep and broken 
country, its cost is probably increased by 50 per cent, 
in extra embankments, retaining walls and extended 
culverts; and the whole question of the cost of laying 
the tramway, compared with cutting metalled road and 
using country carts or a double-railed tram, requires 
careful consideration. 
The laying of the tram in our case was, therefore, 
under the misapprehension that it would work on 
curves of 30 feet radius or less and the cost o£ correct- 
ing this mistake was very considerable. I believe, 
however, that it will work all right when the outrigger 
wheel track is properly metalled, for reasons I will 
presently give, and this I believe is now being done. 
A few of the mistakes made by us are noted below, 
and I think that the ompany did more than its share 
in making Mr. Ewing's invention a practical and useful 
means of transporting material. 
(1) The curves of tlae road were too sharp. 
(2) The trucks were not sufficiently strong. 
(3) The brakes were defective. 
(4) The outrigger wheels were ',too weak (to stand 
bad loading and bad road). 
(5) There was great friction between the flanges and 
the rails. 
(6) There were no proper points or crossings. 
(7) There were no proper lubricating arrangements. 
(8) The rail wheels were oast-iron and they should 
be of good cast-steel not chilled. 
(a) With regard to the sharp curves, we pT3t that 
right, but only at the expense of specially widening the 
ruad and relaying seven miles of rails. 
(h) Trucks are now properly built and two new ones 
can be seen atMsssrs VVal'ijer & Co.'s, Colombo, and 
will, lam sure, interest visitors. 
(c) These trucks have the latest brake designed by 
Mr Michie and myself and it is very efficient. 
Lever brakes, unless they have plenty of travel, 
soon wear the blocks of wood down while screw 
brakes blocked the wheels and caused "flats" to be 
ground on the tread due to friction between rail 
aad wheel, and were therefore unsuitable. 
d. The outrigger wheels supplied were too light 
for the inferior unmetalled road they had to travel 
over and the careless loading; but we have now a 
very good wheel, designed by Mr Kemlo, with tangent 
spokes, giving little trouble. 
As 1 mentioned before, the weight of the loads is 
theoreticalli/ borne entirely by the rail wheels, but too 
often the native truck men load the trucks so that two 
or more cwts of the load is thrown on the out-rigger 
wheel, and besides this, when there is any hollow 
in the road and the out-rigger wheel is below the iiroper 
level, a lot of extra weight is thrown on it ; therefore 
it should be comparatively strong. 
Mr Kemlo has designed a pair of springs, which I 
would have tried and which show exactly the weight 
