October 22, 1909] 



/SCIENCE 



549 



this can be obviated by putting in a ceiling 

 of matched boards after all floors have 

 been laid. It should be finished with shel- 

 lac or coach varnish. Something of a pitchy 

 or resinous nature should be iised (and yet 

 contain no common rosin, as that is far 

 fi-om durable), rather than a paint which 

 can peel or flake off. This should be borne 

 in mind in all overhead construction. As 

 has been said, plaster of any kind is inad- 

 missible in a ceiling on account of its dis- 

 integration by acid fiimes. Cement is no 

 better, as in one laboratory of which the 

 writer knows, a cement ceiling began to 

 flake off within about six weeks after occu- 

 pation. 



Fire Walls and Protectioji.— The build- 

 ing should be subdivided into areas of suit- 

 able size by fire walls extending from top 

 to bottom; all apertures in these walls 

 should be guarded by automatic fire doors. 

 The library, if there be a departmental 

 one, should be housed in a fireproof room 

 and also be protected from being flooded 

 by leaks on the floor above. The more 

 dangerous laboratories— the organic, the 

 oil testing and those below or adjacent to 

 the library should be rendered safer by the 

 installation of sprinklers. Somewhere in 

 the building there should be a fireproof 

 room for the distillation of inflammable 

 substances. In addition to fireproof stair- 

 w'aj's a sufficient quantity of outside iron 

 fii'e escapes should be provided and the ex- 

 its thereto carefully indicated and kept un- 

 fastened. Inch rubber fire hose with nozzle 

 should be provided in each laboratory. 

 Rubber is better than linen or any other 

 collapsible type of hose, in that it does not 

 kink and thus can be taken through door- 

 ways when there are self-closing doors 

 without checking the stream of water. A 

 number of small hose are better than large 

 hose in the hallway, in that they are more 

 accessible, and if used, do not deliver such 



a torrent of water as to occasion a greater 

 loss from water than from the fire itself. 

 Pails of sand with scoops are very efficient 

 and should be found in every laboratory. 

 No money, however, should be wasted on 

 the purchase of "dry powder fire extin- 

 guishers," of which Dr. Freeman says "we 

 recommend that they be thrown into the 

 rubbish heap."- If these are wanted they 

 can be easily made by filling tin tubes with 

 two or three pounds of "anchor dust" or 

 waste bicarbonate of soda. In the organic 

 and oil testing laboratories or any other 

 where the fire risk is unusual, in addition 

 to these safeguards above mentioned and 

 automatic sprinklers, some type of portable 

 chemical fire extinguisher should be in- 

 cluded. This, as is well known, employs 

 carbonic acid generated by the action of 

 sulphuric acid upon baking soda to throw 

 a stream of carbonated water about, which 

 is especially effective in tar and chemical 

 fires. 



A large douche bath with quick opening 

 valve has been found very convenient in ex- 

 tinguishing fire on a student's clothing. 

 This is merely a rose, or better a flat hol- 

 low disk a foot in diameter with concentric 

 slits in it, through which the water issues 

 in a shower; it is set seven feet from the 

 floor. 



Heating and Ventilation. —The so-called 

 "plenum system" for the general heating 

 and ventilating of a laboratory building 

 may be said to work fairly well, but it 

 must be siipplemented by steam radiators 

 and by independent fans, one or more for 

 each laboratory, drawing upon the hoods. 

 These can be placed in the laboratory or, 

 better, on the roof. The hoods in addition 

 to having the usual outlets at the top should 

 be provided with an outlet at the bottom, as 

 most of the gases and vapors of which the 



' " On the Safeguarding of Life in Tlicaters," p. 



