208 
With regard to cows the evidence is not so. clear. Kept shut up 
in “ shippens,” and seldom allowed outside, they have their food 
brought to them, but as this food is composed of vegetable and 
other refuse collected from every possible source and situation, it is 
easy to understand that they can hardly escape from receiving 
infected food from time to time. 
It was interesting to note whether those goats whose blood gave 
a positive agglutination reaction would have some symptoms of 
illness, but this was not apparent except in a few cases. The quan- 
tity and quality of the milk seemed in most cases to be unaffected. 
In fact, it was often noted that the best milk-producers in the herd 
gave a positive reaction. 
Horrocks and Kennedy a thought that as a result of their observa- 
tions, judged by the serum reaction, II per cent of the goats in Malta 
are infected. Ten per cent of the goats supplying milk to various 
parts of Malta appear to excrete the M. melitensis in the milk. 
The excretion of the specific microbe may continue steadily for three 
months without any change occurring in the physical character or 
chemical composition of the 'milk and without the animal exhibiting 
any signs of ill health. Some infected goats may lose flesh and their 
coats become thin ; they may also suffer from a short hacking cough. 
A febrile condition, however, has not been observed. Goats may 
have a marked blood reaction and yet never excrete the J/. meli- 
tensis in the milk. If the blood serum or milk does not agglutinate 
the J/. melitensis , the specific microbe is not found in the milk. 
The excretion of the M. melitensis in the milk may be intermittent, 
appearing for a few days and then disappearing for a week or more. 
A blood reaction may exist for some weeks before the M. melitensis is 
excreted in the milk. 
Monkeys and goats can be infected by feeding with cultures of 
I/, melitensis isolated from milk, and also by feeding with infected 
• milk itself. The incubation period in feeding experiments appears to 
vary between three and four weeks. Monkeys infected by feeding 
sometimes suffer from a typical wave of fever and lose flesh, at other 
times they show no obvious signs of ill health, and may even gain in 
weight. 
When monkeys become infected by feeding with milk the lymphatic 
glands always contain far more colonies of the J/. melitensis than the 
spleen. This fact suggests that the specific micrococci contained in 
the food are carried to the lymphatic glands and there undergo con- 
siderable multiplication. It has not yet been proved that the mesen- 
teric glands are always infected at an earlier date than the femoral 
° Reports of the commission * * * for the investigation of Mediterranean 
fever * * * Part 4, 1906, p. 68, et seq. 
