BUTTER. 301 



througli a fine liair sieve so as to break up curd particles. If 

 too large quantities of starter are used, there is a tendency to 

 produce a sour cheese. The best results are obtained when a 2 

 per cent starter, of the acidity given, is added. 



"In propagating the starter from day to day care must be 

 taken to keep it free from contamination. It should always be 

 prepared in a covered vessel that has previously been sterilized, 

 and tlie milk used should first be pasteurized (or sterilized) and 

 cooled before adding the 'seed.' Some of the original starter 

 should be taken for 'seed,' not the whole milk after the starter 

 has been added. 



''' The starter cannot be used for cheese-making if the milk 

 is overripe, which is the case when the rennet test is 65 

 seconds or under (see p. 282). In sweet milk, testing by 

 the rennet test 120 seconds, the addition of a 2 per cent 

 starter will increase the acidity, so that the rennet test will 

 act in 70 seconds. 



" With sweet milk the use of a pure lactic starter will 

 result in the saving cf 3-5 hours in time. With tainted 

 milk in which the acid develops imperfectly the addition 

 of the starter aids in producing the acidity required for the 

 manufacture of Cheddar cheese." 



BOYD'S PROCESS OF CUEAM RIPEMXG. 



By John Bovd, Chicago, 111. 



It is an accepted fact that the fine aromatic flavor and 

 also the keeping properties of butter depend largely upon 

 the treatment of the cream from the time it is separated 

 from the milk until it is ready for the churn, that is, in 

 the best possible condition to yield the maximum quantity 

 and the best quality as to flavor, texture, solidity, etc., 

 free from casein and other undesirable substances. This 

 perfect condition of cream is understood by tiie term 

 " ripened cream," and when this condition can be pro- 

 duced by the butter-maker with uniformity, regardless of 

 the seasons of the year or extremes of climate, the process 

 may be reckoned as nearly perfect as possible, and not 

 until then. It is most desirable that the process be as sim- 



