166 



These partly round sleepers have all the advantages of square sleepers. The 

 dimensions are increased sufficiently to give an equal strength. The under side is flat, and 

 gives a bearing surface at least equal to, and usually greater than, that of the square 

 sleeper. When the upper surface is round, it is adzed to receive the rails, but too long a 

 bearing is not sought, as it gives sleepers a tendency to rock. The sleepers often contain 

 a large proportion of sapwood, but sapwood impregnated is found to be quite as durable 

 as the heartwood, on account of the greater quantity of preservative substance which it 

 takes up. 



Several important advantages result from the use of such sleepers : 



1. The cost of production is reduced. 



2. The waste in cutting is much less : not 25 per cent., instead of 50 per cent. ; and 

 thus the yield of sleepers may be increased by one-half. 



3. Small trees that cannot give square sleepers may often be cut into partly round 

 sleepers. The Natal Coast forests, particularly those at Umzimkulu, consist mainly of 

 small trees, and would give few square sleepers, but they might be made to yield a larg 

 supply of partly round sleepers. 



Besides Yellowwood, all the other Colonial woods that fulfil the conditions required 

 in a sleeper should be accepted for the purpose. In selecting these woods, the results of 

 mechanical tests given in Appendix II. would be of service. Most of the Colonial woods 

 that are suitable for sleepers are either too hard or too fissile to allow of the driving of 

 spikes without boring, and it would generally be necessary to bore holes for the spikes 

 about half their diameter, and quite through. 



By using sleepers of Yellowwood or of other Native woods with a cellular tissue less 

 compressible than that of Baltic Fir, the use of chairs may be dispensed with, and yet the 

 resistance to the enlargement of the gauge increased. The chairs that are now used cost 

 over a shiling each, and about seventeen hundred thousand will have been laid on the 

 lines now constructed or authorised, if their use is adhered to. 



Metal sleepers have been tried under a variety of shapes, but none have proved entirely 

 successful. They destroy elasticity, and, on account of their deficiency in bulk, do not 

 possess the stability of a wooden sleeper ; they are also more difficult to place and to pack. 

 It does not seem that at present the prospects of the metal sleeper are such as to seriously 

 affect those of the wooden sleeper. 



4. TESTS OF YELLOWWOOD. 



The following extracts give the chief points of information contained in the " Papers 

 and Correspondence tvith reference to the Testing of Yellowwood for Railway Purposes " 

 (Cape Town, 1884) referred to at p. 68. The tests were all performed in England. 



p. 9. Mr. S. Aitken (John Bland & Co., Cardiff), to the Agent-General : " I have 

 now the pleasure to hand you ' Report on ^Native Sleepers ' as experimented upon by me, 

 for the purpose of ascertaining : 



1st. Whether they would take creosote, and 



2nd. Whether in such quantities as would justify my recommending them for 

 Railway work. 



1 am happy to say that I can give you the most positive answer in the affirmative on 

 both points. Indeed, I may say, the capacity of the wood for absorbing the oil has far 

 exeeded my expectations, and I have no hesitation in saying that either of the specimens 

 submitted will make an excellent and lasting sleeper. I consider them superior to any 

 Baltic sleepers you can get, and, if properly operated upon, you have an article in your 

 own hands that will be found efficient and economical, supposing they can be readily and 

 cheaply cut in the forests. Whilst on this part of the matter, I may say, these samples 

 have been submitted to the Great Western Railway Company's inspector, who has had 

 the inspection of their creosoted timber for 30 to 40 years, and also to our own out-door 

 manager, a practical man of similar experience, and they both agree with my estimate of 

 the value of any and all of the sleepers, which they say could not possibly be better, and 

 they regard them as much superior to Baltic redwood. I must now direct your attention 

 to the accompanying statistics, which I have put in simple and complete form, not 

 troubling you with details of the experiments, but simply giving you the results. The 

 five different sorts have each been operated upon in the first instance, all together, so that 

 each has been subjected to precisely the same conditions. The figures require very little 

 explanation, but simply and broadly put, the result is that E, the least absorbent sleeper, 



