[ 54 ] 
within three hours of death, and in two of these within two hours, so that no 
question of post-mortem changes can be entertained. Though these changes 
appear to be common sufficiently so to form an important point in connec¬ 
tion with the post-mortem diagnosis of the disease it does not appear to be 
certain that they are not in themselves secondary. Similar congestive 
changes occur in rabies and in other diseases affecting the nervous system. 
Cultures from the contents of the viscera and from scrapings of the 
congested areas revealed no characteristic organisms, and those from the 
subjacent mucosa, from the glands, spleen, etc., were usually negative or 
showed organisms such as coli which enter the tissues from the intestine 
soon after death. 
Marked and sometimes extreme dilatation of the stomach was a com¬ 
mon factor and occurred both in congested stomachs and in those in which 
there was no trace of congestion. The stomach sometimes contained food 
showing so little signs of digestion that death must have occurred very 
shortly after the food was taken. Considering the cardiac condition it is 
possible that a more restricted or less bulky diet than rice is advisable in 
such cases. 
In other cases the stomach contents were mainly fluid or contained 
food in an advanced condition of digestion. 
The difficulty in deciding as to the real cause of death in some cases is 
great. Absence of patella reflex is not sufficient here to be diagnostic of the 
disease in an acute form. Many of the patients die almost immediately 
after admission and no history is obtainable. In such cases although no 
patella reflex was obtained if, for example, advanced tuberculosis, severe 
acute dysentery, peritonitis or advanced malignant disease is found at the 
autopsy even if Beri-Beri was concurrent the cause of death is not fairly to 
be attributed to that disease. 
On the other hand a patient in the poor condition of health induced by 
chronic Beri-Beri, post Beri-Beric Neuritis, is liable to intercurrent diseases 
which frequently terminate fatally, and in these Beri-Beri is the primary 
cause of death. 
The complications in the cases I have included in my list are as follows:— 
Dysentery, five cases in patients who had Beri-Beri for i to n months. 
Pneumonia, i case in an old Beri-Beri patient whose last relapse was five 
weeks before death. 
Septicaemia, three cases. This is a sequela that is a fairly common 
cause of death. These cases were two of extensive sloughing bedsores, in 
one of which infarcts were found in the spleen, and a third with both 
buttocks gangrenous. 
In these cases death was due to complications rather than to the 
disease itself. 
Two of the patients had old tubercle, but neither extensive nor 
generalised. Two had cirrhosis of the liver. Two had renal calculus. 
In nearly one-third of the deaths regarded as due to dysentery the 
initial diagnosis on admission was of Beri-Beri on the ground mainly of the 
absence of Patella reflex. Whether or not the patients had also Beri-Beri 
in an acute form may be doubted, but the post-mortem lesions of dysentery 
were severe enough in themselves to have been alone the cause of death, and 
I have, therefore, included them in the returns for dysentery, not in those 
of Beri-Beri. In this connection it must be remembered that a form of para 
plegia following an attack of dysentery is not unfrequently met with in 
