The cause of the difficulty has long been charged to various sources,
such as a lack of aeration, improper feeding, retention of animal gases,
etc., but in all these cases it was nothing more than a surmise. Very
often the milk does not betray any visible symptom of fermentation when
received, and the trouble is not to be recognized until the process of
cheese-making is well advanced.

Studies from a biological standpoint have, however, thrown much light on
this troublesome problem; and it is now known that the formation of gas,
either in the curd or after it has been put to press, is due entirely to
the breaking down of certain elements, such as the sugar of milk, due to
the influence of various living germs. This trouble is, then, a type
fermentation, and is, therefore, much more widely distributed than it
would be if it was caused by a single specific organism. These
gas-producing organisms are to be found, sparingly at least, in almost
all milks, but are normally held in check by the ordinary lactic
species. Among them are a large number of the bacteria, although yeasts
and allied germs are often present and are likewise able to set up
fermentative changes of this sort. In these cases the milk-sugar is
decomposed in such a way as to give off CO_{2} and H, and in some cases,
alcohol. Russell and Hastings[210] found a lactose-splitting yeast in a
severe outbreak of gassy cheese in a Swiss factory. In this case the gas
did not develop until the cheese were a few weeks old. In severe cases
the cheese actually cracked to pieces.

According to Guillebeau, a close relation exists between those germs
that are able to produce an infectious inflammation (mastitis) in the
udder of the cow and some forms capable of gas evolution.

If pure cultures of these gas-producing bacteria are added to perfectly
sweet milk, it is possible to artificially produce the conditions in
cheese that so frequently appear in practice.

~Treatment of "pin-holey" curds.~ When this type of fermentation appears
during the manufacture of the cheese, the maker can control it in part
within certain limits. These methods of treatment are, as a rule, purely
mechanical, as when the curds are piled and turned, and subsequently
ground in a curd mill. After the gas has been forced out, the curds are
then put to press and the whole mats into a compact mass.

Another method of treatment based upon bacteriological principles is the
addition of a starter to ind