276 Scientific Proceedings, Royal Dublin Society. 



attention from the worker ; (2) it should operate satisfactorily with air at a 

 constant pressure ; and (3) it should be compact in size. 



General considerations lead to the conclusion that in order to obtain 

 flames of widely different sizes and the same character, with air at a constant 

 low pressure, the air-jet must be variable in bore. For each size of flame 

 there is a particular size of air-jet, which will give a well-proportioned blow- 

 pipe flame with a given air-pressure ; hence one air-jet cannot be made to 

 serve flames of different sizes satisfactorily. With blowpipes operating with 

 a fixed air-jet and foofc-bellows, this defect is partially overcome by varying 

 the air-pressure with the size of the flame, thus supplying somewhat more 

 air to the larger flames by more vigorous working of the bellows. Even this 

 compromise is not quite successful, since, if the jet is of the correct bore for 

 the smaller flames, it cannot supply enough air for the larger ones even with 

 greatly increased pressure. When such a blowpipe is used with a power- 

 blower, however, this difficulty becomes still greater, for it becomes necessary 

 either to vary the speed of the motor for each change in size of flame, or to 

 keep the motor running at the speed required for the largest flame, and vary 

 the air-pressure by means of a tap. 



It was, therefore, taken as a fundamental condition that for a blowpipe 

 intended to work with air at a constant pressure, the air-jet must be variable 

 in bore. The next consideration was to find a way in which this could 

 be achieved without much difficulty. Two methods naturally suggest 

 themselves — (1) to arrange for the actual variation of the air-jet bore by 

 mechanical means ; and (2) to provide a different-sized jet for each size of 

 flame, and to insert each jet when using the corresponding flame. 



The first method is mechanically possible, but it would involve the use 

 of tools and processes which were not at the disposal of the writer, so that 

 attention was mainly directed to the second method. This method is used in 

 an existing type of blowpipe, which is usually called the "glass-blower's 

 pattern " blowpipe, by providing a very wide air-jet, into which short pieces 

 of capillary glass tubing can, with a certain expenditure of time and patience, 

 be wedged with cork or paper to provide the different-sized jets. 



However, it is hardly possible that this device was ever intended to be 

 availed of in the course of work, but was rather intended to alter the blow- 

 pipe when changing from one type of work to another, since the time 

 required to change the jet would be more than sufficient to allow the glass 

 to cool in the course of any one operation. 



'i'he problem then narrowed down to finding a method by which the 

 substitution of different-sized air-jets could be made extremely rapidly, with 

 slight effort on the part of the operator, as the size of the flame varied. If 



