54° 
THF TROPICAL AGRICULTURIST. [February -fc, 1890. 
carriage, to deliver the Bleepers where required at a 
price profitable to both the Railway and Forest 
Departments, aay R3 - 50 per sleeper. It was 
recommended that one-third of the trial sleepers 
in each sample should be used on the sea- 
side and main lines, below Eambukkana ; one- 
third above Eambukkana, on the Matale branch and 
as far as Nawalapitiya ; and the remainder above 
Nawalapitiya. They will thus be tested in all the 
conditions of soil and climate, ballast and traffic 
likely to occur in the colony. The sample sleepers 
will, of course, be carefully and conspicuously 
marked, with name of species and year of felling, 
so that no mistake as to the identity of any can 
occur ; and with constant examinations and 
reports by the Railway Engineer, it is anticipated 
that even in one year valuable results may be 
obtained, the minimum final period of trial beiDg 
three years. Should the reports be favourable and 
the timber abundant in any limited locality it is 
believed that the purchase and transport of steam- 
sawing machinery would be advisable. Several 
witnesses gave evidence to the effect that thus a 
saving of 50 per cent might be effected in the sawing. 
Mr. Huddleston, the Assistant Forester of the North- 
Central Province, suggested the use of wind 
power at Trincomalee, and we think the idea 
well worthy of attention. The Committee also 
suggested the purchase for the use of the 
Forest Department of a small elephant es- 
tablishment to drag timber to the sawing mill. 
Hired elephants, even in this land famous for its 
elephants, are difficult to obtain and during festivals 
unobtainable. 
Such is the general tenor of the Report, founded 
on interesting and valuable evidence, some points in 
which we shall quote and notice in a further article. 
Meantime, we cannot help saying that the very 
difficulty of obtaining full supplies of Bleepers for our 
existing railways, owing to the remoteness of forests 
which teem with suitable timbers, forms the 
strongest argument for the extension of railway 
facilities to as many as possible of such remote 
localities, or to the nearest proximity to them, 
without too severe an application of the 
"Will it pay?" principle. Such Railways would 
while opening up the country to enterprise and indue- 
try, gradually develop traffic which would enable them 
to be directly profitable ultimately ; while meantime 
they would help largely to make existing lines pay 
better than they now do. 
We now come to the evidence given before the 
Committee on railway sleepers, and the results of 
Mr. Cantrell's extended experience in regard to 
experiments wilh sleepers, imported and of local 
timber, are so important that we feel bound to 
give his evidence in full, except that in regard to 
the Kjrri sleepers, the pu port oi which has already 
been ttated. Mr. Cantrell stated : — 
DuriDg my 22 years experience of Railway work 
in Ceylon I have tried several kinds of sleepers, 
imported, and native. Th- imported ones are creosi ted 
Baluc fir from EDgland, red and blue gam sleepprs 
and Jarrah for bridge limber from Australia and 
Johore Teak. Of native woo s I l.ave had ex- 
per erice of Satinwood, Mililla, Kahata, Na, 
Palu, Mi, Red doon and White doon. Other 
kinds were tiied by Mr. Faviell, the contractor for 
the main line viz. Del, Caroo, Damhu and Kura, but of 
these I have no record , they were all tried in the low- 
country. With regard to creosoted pine sleepers they 
have. U'O merit of b> ing choaper than any other sleeper 
in first cost. This is important only on the straight and 
level portions of " the road," but on sharp curves aud 
steep gradients, their life is short and they are com- 
paratively more expansive than others ; on sharp curves 
I should use nothing but hardwood sleepers if we could 
get a sufficient supply, but as the supply is short, we 
use the two kinds alternately. My experience of the 
sleepers laid on the Nanuoya section of the line is as 
follows : — 
On sharp curves under 7 chains the road was laid al- 
ternately with creosoted and hardwood sleepers, the 
hardwood sleepers consisted of red doon, white doon, 
kahata, and mi. These sleepers were not applied under 
Government supervision, but were purchased from 
planters and others and it is impossible to say whether 
the woods supplied were really what they professed to 
be. Experience shows that these woods were not of 
good quality for sleepers with the exception of the red 
doon. The average life of a creosoted sleeper maybe 
taken at 8 years throughout the line ; the average life 
of a satiuwood, or mililla sleeper, 12 or 15 years, of red 
gum from 12 to 14 years, of kahata not more than 5 
years, of white doon less than 5 years and of red doon 
about 8 years. Na is not a good wood for sleepers and 
the same remark applies to palu ; but I must qualify 
this statement as ie.j~.rd8 palu by saying that the 
shipment of about 5,000 which came from Jaffna con- 
sisted manifestly of immature trees. Creosoted 
sleepers generally fail from getting too thin at 
the rail bed owing to the sleeper being cut into by 
the rail. The ends are very apt to split and to break 
off when in this condition. The crushing strain and 
the frequent respikiDg in these soft woods, in order to 
keep the rails to gauge, wear the sleeper through. The 
conditions on the sea side line are in my opinion more 
unfavorable to oreosoted pine sleepers than on the 
inland sections so that I do not estimate the life of such 
sleepers on the sea side at more than 6 to 7 years, but 
including the inland sections the average life may be 
taken at 8. The life of a Johore teak sleeper is 10 years. 
Its chief defect is its liability to split, f jr which reason 
it requires to be bolted or ringed at the ends. This is 
the experience of 10,000 of these sleepers and they were 
exclusively laid on the section between Colombo and 
Kandy. Their cost in 1876 was R5 each. 1 prefer 
satin or mililla wood to Johore teak. Jarrah has never 
been used for sleepers in the road, but it has been tried 
for longitudinal bridge timbers. It has a life of 12 years. 
Its defect is liability to dry rot. Mi and nearly all other 
native wood sleepers that have been tried excepting 
satin and mililla fail eventually from dry rot. Mililla, 
satin and both the red and blue gum sleepers failed 
eventually from old age. The cost of a creosoted sleeper 
varies from R2'45 to R3'75. If I could get a suitable 
native hardwood sleeper at a reasonable price I would 
use it on the curves exclusively. By a reasonable price 
I mean R5 for a sleeper which would give an average 
life of 10 years or upwards. We require from iO to 
30,000 sleepers per annum to maintain those parts of 
the line at present open to traffic and, of these, we could 
use as many as 15,000 hardwood sleepers. 
Due weight will be given to the qualifications stated 
by Mr. Cantrell, that the sleepers of indigenous 
wood used by him were in many cases not care- 
fully identified as to species of trees, or tested 
as to age ano quality of timb r, so that it is ri«ht that 
Na, so h ghly spoken of by many, and other timbers, 
said to ho.ve failed as sleepers, should have a furtner 
trial. It seems a great pity ihat the remits of 
Mr. Faviell's experience should not be available, 
but we should think Mr. Kendrew must have 
kept records which could be referred to. The 
next witness examined was Mr- A. F. Broun, 
Deputy Conservator of Forests. In answer to the 
question whether the forests of Oey.on were ade- 
quate to the supply of sleepers for the railway 
he said : — 
As far as my Indian experience goes I have no doubt 
on the point. In support of this view I may state that 
a tree of 6 feet girth breast high, and I am taking 
a low fstimate, yields at least ten sleepers. Tbat means 
that 1,500 tree- altogether would h ve o 1 e torn she-i 
by the whole i-laud, or roughly 200 tree- per Province. 
A great many indigenous tree* I think would give 
sleepers at least as good as the creosoted pine, which 
is still being used on the Bailway. Lurge trees of 
Palu for instance would often yield 120 sleepers, In 
