REVIEWS ELEMENTS OF INORGANIC CHEMISTRY. 491 



dipping a little into the liquid. If we call the flask A, and the bottles 

 Bj C and D, then the straight tuhe in C prevents the reflux of the 

 liquid from D, caused by a partial vacuum or absorption in C ; the 

 tube in B does the same for C, and the bent tube in A allows air to 

 enter if a vacuum be produced in the flask and the wash water rises 

 into the delivery tube, thus preventing reflux from B. 



In the work before us, and in many others it is recommended, when 

 we have no three-necked bottles, but only those with two apertures, or 

 when we are compelled to use a cork with two perforations fitted into 

 a wide-mouthed vessel, to insert into one neck or perforation a wide 

 tube dipping into the water, and to pass the gas-delivery tube through 

 this into the liquid. If the above series were arranged in this way, 

 then the wide tube in C would prevent reflux from D, in B from C, 

 but the wide tube in B would have no action in preventing reflux from 

 B into A, that would have to be guarded against b} a safety tube 

 attached to the flask itself. And yet we find such an arrangement 

 recommended in Graham's Elements, which by its adoption might 

 frequently lead to very disagreeable if not dangerous consequences, 

 and against which the beginner in practical chemistry cannot be sufli- 

 ciently warned. 



At page 275 we have an apparatus for evolving carbonic oxide from 

 a heated mixture of sulphuric acid and ferrocyanide of potassium 

 (Fownes' process). The flask is without safety tube, but the gas tube 

 passes through a wide one into the washing bottle, which it is clear 

 would not prevent the possibility of the wash water running back into 

 the hot sulphuric acid. 



At page 28? we have a similar want of safety tube in the preparation 

 of olefiant gas. 



At page 294, fig. 138, the same contrivance is employed in the ap- 

 paratus for preparing a solution of sulphurous acid, and in the text 

 we read that the washing phial and wide tube serves to " prevent the 

 liquid in the second bottle from passing back into the generating flask, 

 on the occurrence of a contraction of the air in the flask, by cooling 

 or any other cause." It undoubtedly would prevent the return of the 

 water from the second bottle into the first, but not of that in the first 

 into the flask, and by the neglect of the proper precautions in this 

 process very dangerous explosions may sometimes happen, as the 

 writer knows to his cost. 



