154 



days The fermentation of the liquor is a very quiet one, the gas being 

 given off slowly in small bubbles, and at some stages fermentation is 

 hardly noticeable. During fermentation, although there is a somewhat 

 "fruity" smell, a definite aroma cannot be said to be produced, but after 

 the fermentation is concluded, if the liquor be allowed to remain quiet, 

 say from 24 to 36 hrs. a delicious aroma can be distinguished. In order to 

 prove without a doubt, that the aroma produced was due to this germ, 

 the following experiment was performed. A certain quantity of molas- 

 ses and dunder and water, mixed together in suitable proportions, was 

 taken and sterilised by boiling. It was then allowed to cool in contact 

 with air previously freed from all germs, and when a sufficient amount of 

 air had been absorbed the liquid was equally divided between two fer- 

 menting cylinders which had also been previously sterilised. One cylinder 

 was set in fermentation by means of this particular yeast, which I call 

 No. 18, and the other cylinder was fermented by another Jamaica yeast 

 which I will call No. 4. The two cylinders were then placed under ex- 

 actly the same external conditions, and fermentation allowed to proceed. 

 The appearance of the two cylinders during fermentation was charac- 

 teristic. No. 18 was covered with a thick golden buttery head, and fer- 

 mentation was slow, while in No. 4 cylinder the yeast remained entirely 

 at the bottom, and the fermentation was rapid, and was what is techni- 

 cally called a " champagne" fermentation. At the end of 5 days fer- 

 mentation was entirely at an end in No. 4 cylinder, while in No. 18 it 

 was still in progress. No 4 cylinder was allowed to stand 36 hrs. 

 No aroma was developed. Eventually fermentation was finished in 

 No. 18 in 12 days, and the wash allowed to stand 36 hrs., — a heavy fruity 

 aroma teas developed. This experiment clearly shews that the aroma in 

 question was due to the influence of No. 18 yeast, since the two washes 

 fermented were identical in composition, and were fermented under ex- 

 actly the same external conditions. It raises too a point of some 

 practical importance to which I would call the attention of estate 

 owners and distillers. There seems to be a general unanimity of opinion 

 among planters that in order to produce a fine rum, the " wash" 

 must be allowed to <: die down" thoroughly. To accomplish this however 

 necessitates in many cases building larger still houses, which many 

 estates in these hard times are unable to do. But does it not seem evi- 

 dent from my experiment that the amount of benefit to be derived from 

 the enlarging of the still house and thus giving the liquor room and 

 time to attenuate thoroughly will vary very much according to ivhether 

 an aroma or non-aro?na producing yeast has the mastery in the vats ? So 

 far we see the problem must be approached from two sides. The case 

 however presents other points of interest. How far in the experiment 

 under discussion was the aroma due to No. 18 yeast ? Did the yeast excrete 

 the aroma, so to speak, or did it form it from, or by transforming, certain 

 substances in the liquid ? In order to settle this question, I fermented 

 separately, by mean s of No. 1 8 yeast, refined cane sugar, dextrose, cane juice, 

 and molasses, leaving the liquor to stand 36 hours after the completion 

 of fermentation : in no case was the aroma developed. I repeated these 

 experiments, but the results were the same. I need hardly add, that 

 where necessary, yeast nutriment was added in order to produce a normal 

 fermentation. This puzzled me for some time. Dunder and molasses and 

 No. 18 yeast gave the aroma, but molasses and nutrient salts and No. 18 



