1896.1 MICROSCOPICAL JOURNAL. 321 



to oxvi^en, oxide of ni;ing-ancse and lime. Then, to carry 

 off the surplus of permang-anate and complete the ])uriti- 

 cation, the water is poured over bioxide of man<>"anese; 

 oxyg"en in the nascent state is thus freed and it burns up 

 any remaining- g-erms. There remains in the api)aratus, 

 then, inferior oxides of mang-anese, which hasten to reoxi- 

 dize themselves and furnish ag-aiii a certain quantity of 

 bioxide of mang-anese ; the water, as thus finally purified, 

 contains a little lime in the form of a bicarbonate and 

 traces of oxyg-enated water. A very small quantity of 

 permang-anate of lime is used in this process, and, if prac- 

 ticable on a larg-e scale, is of g-reat importance. Water 

 having- 100,000 colonies of microbes can thus be purified, 

 it is stated, and ice placed in water with permang-anate of 

 lime is also quickly sterlized. — Sanitarian. 



Enzym in Malt. — Linter observed that dextrose was 

 formed by the action of malt extract or precipitated dias- 

 tase on starch. As Morris has denied the presence of g-lu- 

 cose in malt, the author undertook an investig-ation to de- 

 termine the presence of a dextrose-forming- enzym in malt 

 and the conditions under which it acts. The results were 

 as follows : 



(1) Malt contains dextrose, sucrose, probably levulose, 

 but no maltose. 



(2) The absolute and ralative amounts of dextrose and 

 sucrose are very variable. 



(5) In malt extracts (prepared at 15 deg-rees and 55 de- 

 g-rees) no. ferment which inverts sucrose was found. 



(4) Malt contains a dextrose-forming ferment which 

 seems to act most energ-etically at 55 deg-rees. 



(5) Roasting- chang-es the reducing- sug-ars in malt to 

 products having- a smaller reducing- power. — Expcrimetit Sta- 

 tion Record. 



On the Enzyma of Some Yeasts. — The bottom yeasts 

 (type Frohberg- and Saag-) contain an enzym which breaks 

 up melibiose while the surface yeasts of the same type 

 have no appreciable action. As the latter contains consid- 

 erable invertin, this result was a direct contradiction of 



