4*6 
favoured particularly by them in this way is not many yards away 
from the unprotected larder window of one of the hotels. This 
instance is only quoted as one among a large number. 
T emperature seems also to have a very marked effect upon the 
frequency of the disorder. In Las Palmas, for instance, the summer 
and winter temperatures vary only by about io degrees, and the 
temperature does not usually go below about 75°F., so that any 
bacteria carried on to food-stuff by flies would be under very 
favourable conditions for rapid multiplication. In the Monte district 
oi the same island the temperature is much more variable, and during 
the winter is comparatively low. In this district cases of so-called 
Canary fever are comparatively rare in the hotels. 
These observations indicate that there should be a greater risk of 
conveying a large dose of bacteria, and of the toxins they produce, by 
means of soup from the stock-pot and of rechauffes, than by any other 
kind of food. The soup is rarely if ever brought to boiling point, and 
necessarily is allowed to cool every night when the kitchen fire goes 
out. Entrees and other forms of rechauffes are made from meat or 
is that has been previously cooked, and such material will have been 
e t or from 18 to 24 hours before being served without having been 
rought up to the boiling point of water. Micro-organisms will have 
een destroyed, and many of their toxic products may have been 
1 n ^ e fc> ra ^ ec ^ by the temperature generally required to cook food for 
rst time, and with such material the risk of infection or 
ft 1. 1U1 ^ muc h less than in the case of the soup from the 
Suffir' °, r .°[ 3 1 ec bauffe which is rarely if ever brought up to a 
sufficiently high temperature. ' 
hotel 1 - T St l ' H1 P ' as ^ >a ^ mas . most of the other visitors in the 
avoided attacks °f the disorder. My two companions and I 
occasion ?° UPS anc * entr ^ es » and also avoided infection. On one 
followed w>r eVer ’f° nC ° f my com P anions tQ ok an entree. An attack 
disease is re e, " a ' en ,n conjunction with the fact that the 
in the method oTl- !° ^ !'° tGls ’ and kee P in g in mind the difference 
-girded as corroborate evidence^ ***** ^ *° ^ 
being confined^ hL ^ f, ' S ° rder ’ its epidemic character, and its 
hotels, suggest very strongly that it is due to 
