made of wax, carefully moulded, and their weight is automatically registered. There is an over-head rail for removing the models from one place to another, while the carriage from which the model is towed through the water runs on rails fixed on each side of the concrete walls of the tank, and is driven by electricity. At about the centre of the main tank building there is an observation room which is used for photographic purposes. Messrs. John Brown and Co. themselves have admitted that it is owing to the valuable experiments obtained in this tank that they have been able to design ships producing the best results, whilst also exhibiting the maximum economy. Mathematical theories and formulæ have contributed much to the development of the steamship, but there is a point reached when these are of no avail for the reason that when new problems arise that cannot be solved by former experiences and existing data, a more practical method of obtaining information must be found. It is here that the tank comes in to solve the difficulties at hand both as to the hulls of the ships themselves and the character of the propellers which are to send them through the water. Had the experimental tank been encouraged at an earlier date, no doubt certain of the errors which characterised some of the ships of the sea might have been avoided. It is not enough to build a steamship of enduring strength, and to give her the best engines of the time; it is also essential that she be designed in such a manner that her propellers forge her ahead with the minimum of resistance. Germany and America, no less than Great Britain, are now busying themselves with the employment of the naval experimental tank, and obtain thereby so many valuable data as to make such institutions indispensable if advance in the science of naval architecture is to be something more than ephemeral. The Norddeutscher Lloyd Company had such a tank built in 1900 on the model of the one belonging to the Royal Italian Navy at Spezia, and some description may not be without interest. The tank is contained in a building 170 metres long and 8 metres wide. On either side of the tank is a strong set of rails on which the towing carriage runs, and the building contains workshops wherein the models are constructed. The experiments are not complicated, for after the displacement of the projected ship has been decided on, several models of such a displacement are made from drawings by means of an ingenious machine. These models are made out of paraffin wax, and about 4 or 5 metres long. (A metre, it should be remembered, is the equivalent of 1·094 English yards.)