129 



inexpensive one. The points to be secured are : first, 

 a very thin layer of the liquid under examination, and, 

 second, the coagulation of the albuminous constituents 

 of that liquid. A drop, a very small drop of pus, sputum, 

 or blood is placed in the middle of a clean cover-glass ; 

 a second cover-glass is then laid upon this and the two 

 are pressed gently together until the enclosed drop has 

 become a very thin layer ; the two cover-glasses are then 

 separated by sliding one over the other, not by pulling 

 them apart. The two glasses, each of which has now a 

 thin layer upon its surface, are allowed to dry,; the co- 

 agulation of the albumen of the fluid (a most important 

 measure, which prevents precipitation of granules from 

 the staining fluid) is now accomplished by holding the 

 cover-glass over a gas or alcohol flame until it acquires 

 such a temperature as to be unpleasantly warm when 

 applied to the skin ; or the glass may be " passed three 

 times through the Bunsen flame," as directed by Ehrlich. 

 The pus or sputum is now stained either by floating the 

 cover-glass on the staining liquid, or (more conveniently) 

 by pouring a few drops of the aniline solution from a 

 pipette onto the cover-glass ; after five to fifteen minutes 

 the staining fluid is poured off, the excess of color 

 removed by washing in a stream of water ; the cover- 

 glass is now allowed to dry and is then ready for mount- 

 ing in either glycerine or balsam. The methylene blue 

 is perhaps the most satisfactory of the aniline colors 

 for general use in staining these dry cover-glass prep- 

 arations. 



While most species of bacteria known to occur in 

 human tissues are readily stained by these methods, 

 certain varieties require special modifications thereof. 

 Piominent among these is the bacillus found in tubercu- 

 lous tissues, whose presence in sputum already possesses 

 diagnostic (and prognostic ?) value. Of the various 

 methods already published for staining the bacillus tu- 

 berculosis, Ehrlich's has the general preference, and has 

 always given satisfactory results in the writer's hands. 

 This method requires (a) a saturated solution of aniline 



