51 
It has been a matter of special interest to me to compare this 
hypothetical scheme of generations with pellagra data communi¬ 
cated to me by Dr. H. Douglas Singer, in a letter written Decem¬ 
ber 29, 1911. The statements of this'letter are illustrated by a 
curve showing the number of fresh cases of pellagra occurring at 
the Bartonville general hospital for each month from July, 1909, 
to September, 1911. There are five high points in this curve for 
these twenty-six months—one at the beginning of the record, 
which, starting with 21 cases for July, 1909, rises to 71 for Au¬ 
gust, and then drops rapidly away to 3 in December. This is much 
the highest wave of the curve. The next wave of increase begins 
with a single case in April, 1910, rises to 16 cases in May and to 
34 in June, and drops to 4 in July. We have next a lower wave 
of 16 cases in August, 1910, 15 in September, and 1 in October. 
Two cases in the following March (1911) become 4 in April and 
6 in May, fall to 1 in June, rise again to 3 in July and 7 in 
August, and fall to none in September. 
On the supposition of a connection between black-fly outbreaks 
and pellagra waves, we should naturally expect the former to pre¬ 
cede the latter somewhat—just how far, of course, no one can tell, 
since that would involve a knowledge of the incubation period of 
the disease. A comparison of my hypothetical periods of our 
black-fly generations with these waves of frequency of new cases 
of pellagra, gives an indication of a correspondence between the 
two series for the first two generations of the year, but negatives 
the idea of any stimulating influence of the implied third genera¬ 
tion. Thus, omitting the 1909 record as begun too late to serve 
our purpose, the March and April generation of .black-flies for 
1910 connects with an April and May increase of pellagra; the 
supposed June and July generation of that year, with high num¬ 
bers of new cases for August and September; the March and 
April generation of 1911 connects with an April and May increase 
of pellagra for that year; and the June and July generation with 
a July and August increase. The August to October generation, on 
the other hand, is followed by a decline in the number of new cases 
in both 1909 and 1910, the record for 1911 breaking off within 
this period. 
These interpretations, it is true, are decidedly hypothetical, but 
they may be taken as at least suggestive of a causal relationship, 
and as indicative of a method of analysis which, used in proper 
cases, may give definite results. We need to know accurately the 
life histories of the various species of Simulium for the entire 
year in some locality where pellagra is more or less prevalent, and 
to know also the exact facts as to the local abundance of the winged 
