518 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



with a block of wood shaped to the curve to which it is de- 

 sired to bend the pipe, its edge being grooved to fit the out- 

 side of the pipe. After the bend is made the mandril can 

 readily be withdrawn by turning it from one end in the di- 

 rection in which it is wound, which action slightly reduces 

 its diameter, and by this means it is screwed out very easily. 

 The old method of bending pipes of copper or brass is to fill 

 them first with rosin, lead, or sand, and requires considerable 

 skill, being both tedious and uncertain, especially for large 

 pipes; and the crimping of the inner sides and stretching of 

 the outer sides are very unequal, the pipe always requiring 

 to be hammered and filed into shape at the conclusion of the 

 bending. The new method, besides requiring much less skill 

 and scarcely one twentieth of the time of the old, obviates 

 all the above-named objections, leaving the pipe so near per- 

 fectly round that the eye does not detect the variation, and 

 its surface so smooth as not to necessitate filing. 1 -^, 

 LXX., 22. 



TEMPERED STEEL BRUSHES. 



A recent novelty in the metal trade has been introduced 

 by a French manufacturer, in the shape of brushes made of 

 tempered steel wire. These implements, according as they 

 are intended for various uses cleaning tubes, or the sheath- 

 ing of ships, iron bridges, sluice-gates, and iron construction 

 generally are of different forms. They are also made, like 

 the old metallic brushes, for polishing or scratching metals, 

 for stone and marble, for smoothing flooring and the decks 

 of ships, and for many other purposes. The same maker has 

 produced a new scraper for the special purpose of cleaning 

 locomotive boilers. This special device is a spiral brush, in 

 which, however, the steel wire is replaced by pieces of steel 

 spring, and is said to be not only very effective, but also very 

 durable. Iron Age, September 2, 1875, 1. 



RAPIDITY OF FILTRATION. 



Dr. Fleitmann has called attention to the fact not gener- 

 ally known that, contrary to what at first sight might be 

 expected, filtration is much more rapid through thick paper 

 than through thin, and that it is almost twice as rapid through 

 a double filter as through a single one, and still more rapid 



i 



