1894.] MICROSCOPICAL JOURNAL. 39 



the thermostat by 25°C. Afterwards they are examined 

 every day. 



We now return to the case in question. We had 32 

 colonies each of which was transferred into a flask. Two 

 days after this infection there was a vigorous growth in 

 all of the flasks, and, when the microscopic pictures was 

 compared, there were three different forms of cells rep-, 

 resented : 



1. Round and oval isolated cells in flasks No. 1, 3, 4, 

 10, 12, 15, 18, 25, 27, 29 (see Fig. 3). 



2. Round, oval and pastorian cells in flasks No. 2, 5, 

 6, 7, 8, 9, 22, 24, 28, 30, 31, 32 (see Fig. 4.). 



3. Round and oval cells in short rows in flasks No. 11, 

 13, 14, 16, 21, 26 (see Fig. 5). 



Flasks No. 17, 19, 20, 23 gave a mixed picture, and 

 were, therefore, discarded. 



From each of these three series, two of the most typi- 

 cal ones was selected and " freshened up "" in new wort 

 for 24 hours three times. After this they were trans- 

 ferred to gypsum and their specific natures determined. 

 No. 1 was S. cerevisice I (var. easily forming spores), 

 No. 2 S. Pastorianus III, and No. 3 S. cerev. I (var. not 

 easily forming spores).* 



If an infection by moulds or spores of these is sus- 

 pected, a wort-gelatine plate should be prepared. Here 

 the colonies of cultivated yeast are round or slightly 

 oval, while the colonies of wild yeasts (especially S. Pas- 

 torianus III), have fringed edges or, when the edges are 

 even, they are somewhat cone-shaped. Colonies of 8. 

 tnycoderma {M. cerevis'rx) are deepened in the middle, 

 and thus easily recognized. The moulds {Mucor, Euro- 

 Hum, Penicilliwrn) appear with their characteristic my- 

 celium and fructification, and cannot be mistaken. The 

 preparation of gelatine plates is described in many bac- 



*Since systematic data were given in my paper (Am. Nat. XXVII) they 

 are not repeated on this occasion. 



