IQO 



BACTERIA IN RELATION TO PLANT DISEASES. 



to one or more parts of a large drop of water on the cover, which is then put into 

 the thermostat, so that it sliall dr}- quickly and yet allow time for various rods to 

 swim free from the tangle of their fellows ; if the mordanting is thorough, but not 

 excessive, and, iinally, if there is no unforeseen mishap in the subsequent staining, 

 good preparations of man}' organisms are easil}' secured. Others ha\'e given the 

 writer much trouble {Bacillus amylovoriis is one of the worst), and the only con- 

 clusion he has been able to reach is that the bacteria vary greath- in their response 

 to flagella stains. Sometimes well-cleaned covers give trouble and then the surface 

 of the glass itself is at fault (V. A. Moore), and covers of another origin should 

 be tried. 



Young agar cultures are usually preferred, /. e. , those not over twenty -four hours 

 old, but good results may occasionally be obtained from much older cultures and 

 also from other media. In case flagella are to be stained from fluid cultures by 

 ordinary methods, the medium in which the organism is grown must be vei"\- dilute^ 

 and there must usually be an additional extensive dilution on the cover, or before 

 putting thereon, to avoid a dense ground stain. Bouillon contains too many fine 

 particles, but Zettnow has found a way to stain from bouillon (Bibliog., XII). 



Recently the writer has obtained ven? good results with several organisms b)- 

 growing them for some days in lo cc. of distilled water to which 2 or 3 drops of 

 Uschinsky's solution was added. Covers were prepared directly from this solution. 

 Others have reported good success by transfer of agar-growu organisms to consider- 

 able water in a watch-glass or test-tube and incubating this in the thermostat for 

 some hours before preparation of the covers. Others have recommended the use of 

 filtered, sterile, h}'drant water instead of distilled water. Nearh' ever)' worker has 

 some favorite stain. The writer prefers Van Ermengem's. 



Locfflcr's Flagella Stain. 



( 1 ) Mordant : 



Solution of tannin (20 per cent in water) .... 10 

 Saturated (cold) aqueous solution of ferrous 



sulphate* 5 



Saturated alcoholic solution of hasic fuchsin. i 



(2) Stain: Carhol-fuchsin. 



(3) Correcti\'C solutions: (a) i per cent solu- 

 tion of caustic soda; (b) a sulphuric-acid solu- 

 tion of ecjuivalent strength. 



f'. .■/. Moore's I'lagclla Stain. 



This is a slight modification of LoefBer's. 

 d) Jlordant: 



Tannic acid (20 per cent in water ) 10 



Ferrous sulphate (cold saturated water solu- 

 tion ) 5 



Basic fuchsin (saturated alcoholic solution). T 

 (21 Stain: Zielil's carlml fuchsin. 

 LIse the mordant frcsli, and filter each lime 



before UMiig. Fix tlie him on the cover-glass 



by passing it quickly face up twice througli the 

 open flame, or by exposing for 5 or 10 minutes 

 lo a temperature of 120° to 140° C, the better 

 way. A few cubic centimeters of the mordant 

 are now placed in a wide test-tube (i inch), the 

 cover is lowered into it, and the mordant is 

 heated over a flame until steam rises. It is 

 then removed from the flame and the cover is 

 allowed to remain in the hot fluid 5 or 10 

 minutes, with occasion.al gentle shaking, after 

 which it is drawn to the mouth of the tilted 

 lube by means of a hooked platinum wire set 

 into a glass rod. After thoroughly rinsing under 

 the tap or in a hue stream of water the cover 

 is lowered into the stain (held in another wide 

 test-tube), where it is heated until steam rises 

 (I to 3 minutes). One should know from be- 

 ginning to end which side of cover bears tlie 

 bacterial film. The cover is drawn to month of 

 tube by means cd" a hooked platinum wire. Some- 

 limes 1 per cent sol. NaOH ('j cc.) may l.)e 

 :idde(l to mordant with advantage. 



*Tlie iron oxide may be nmovcd from tlie solution of ferrous sulpluite by p.issiiiK it tlirouRh a filter pajicr. 

 Selected cr>-stals slioidd be used. 



