ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 571 



Mix A and B, shake and filter. Add to the filtrate 40 com. of hydrogen 

 peroxide. Shake again. The fluid undergoes the changes described 

 above, and becomes at once ready for use. 



Pikrin Method of Staining Tubercle Bacilli.* — Spengler's method 

 consists of the following steps : — 



1. Stain with warm carbol-fuchsin. 



2. Pour off the stain without washing. 



3. Pour on picric acid alcohol (equal parts of saturated solution of 



picric acid and absolute alcohol). 



4. After 3 seconds wash with 60 p.c. alcohol. 



5. Treat with 15 p.c. nitric acid until yellow (30 seconds). 



6. Wash again with 60 p.c. alcohol. 



7. Counterstain with picric acid alcohol until lemon-coloured. 



8. Wash with distilled water. 



H. Wilson finds this method far superior to that of Ziehl-Neelsen, 

 especially when applied to the examination of urinary sediments. 



Modern Methods of Sputum Investig-ation.t — Gr. H. K. MacaUster 

 gives a statistical analysis of the results obtained at the Lister Institute 

 with the use of autiformin in the routine examination of samples of 

 sputum for the presence of tubercle bacilli. Out of 2273 samples, 622 

 were found upon simple direct examination to be positive. The remain- 

 ing 1651 specimens, which gave negative results, were incubated over- 

 night with an equal bulk of 30 p.c. autiformin, and centrifugalized. 

 The sediment was washed with distilled water. Examination then 

 showed the presence of tubercle bacilli in 9 cases. So that of the sputa 

 found negative upon direct examination, only 0*54 p.c. proved positive 

 after antiformin treatment. This improvement percentage of " 54 p.c. 

 contrasted with figures ranging from 10 to 35 p.c. recorded by other 

 observers, is explained by the fact that considerable time is devoted to 

 the direct examination of the original sample. 



A comparison of the Ziehl-Neelsen staining method with those of 

 Herman, Much, and Gasis, shows that for sputum examination the first- 

 named remains the most reliable and satisfactory, while the relative 

 degrees of merit of the other procedures are shown by the order in 

 which their names stand. 



Haematoxylin as a Bacterial Stain. J — A. Feeser has investigated 

 the value of hematoxylin for the staining of bacteria. He has found 

 that, in spite of statements to the contrary in text-books, this stain is 

 under certain circumstances quite suitable for bacterial preparations. 

 Bohmer's stain, freshly prepared, only stains bacteria after several hours' 

 action, but if matured by three months' keeping, it stains ordinary 

 bacteria well in an hour, faintly in 15 minutes. If heat be applied, 



* Brit. Med. Journ., 1912, ii. p. 413. 



t Brit. Med. Journ., 1912, ii. pp. 411-13. 



j Centralbl. Bakt., Ite Abt. Orig., Ixvi. (1912) pp. 137 42. 



