424 
bulletin of the bureau of fisheries. 
McCarrison, in his observations (1906) on endemic goiter in India, considers the 
dissolved content of the waters concerned and is unable to find anything of significance 
in its relation to the amount of goiter. For the villages of the Chitral Valley, all centers 
of goiter, he gives the foregoing table (table iii). The meager chemical data here 
shown afford no clue to the explanation of the distribution of goiter in the villages. 
He gives similar data for the single water supply of the several component villages 
known collectively as Gilgit, which are located in a series along the water channel from 
above downward, so that an increasing pollution occurs toward the lowermost Aullage. 
There is here an increase of the incidence of goiter from above downward, culminating 
at the lowermost village. In the following analysis of the Gilgit water supply, the 
Barmis water is an accessory supply which drains into the main channel. No case of 
goiter has been found among those who use exclusively this Barmis water. 
Analysis of Gilgit Waters. 
^Supply. 
Total 
solids. 
Parts per 
100.000, 
Total 
hardness. 
Grams to 
gallon. 
Calcium. 
Grams to 
gallon. 
Magne- 
sium. 
Grams to 
gallon. 
Iron. 
Grams to 
gallon. 
Sul- 
phates. 
Grams to 
gallon. 
Chlo- 
rides. 
Grams to 
gallon. 
Free. 
NHj. 
Organic 
matter. 
Other 
metals, 
lead, 
copper, 
zinc. 
Gilgit 
“30 
7 - 143 
6 
Trace. 
Trace. 
3 
Nil. 
Nil. 
Nil. 
Barmi'^ 
b 19 
9 
8-10 
Trace. 
Trace. 
3 or over. 
Nil. 
Nil. 
Nil. 
Nil. 
21 grams to gallon. 6 13.3 grams to gallon. 
It is thus seen that McCarrison’s chemical evidence is negative in its bearing on the 
origin of goiter. The analyses are far from exhaustive, but it is to be doubted whether 
greater detail would be more significant. The much more complete data which we 
show for waters associated with thyroid disease in fishes is as fruitless of any theory of 
causation. 
ENDEMIC OCCURRENCE. 
C.ALEDONIA HATCHERY, N. Y. 
Studied from the standpoint of visible tumors, we found that in certain hatcheries 
the disease is endemic. It is clear that a large proportion of the hatcheries which main- 
tain Salmonidae under conditions of domestication are more or less affected by thyroid 
tumors. Studied from the standpoint of visible tumors, the history of the disease in a 
given hatchery may be traced backward in some cases for seA'eral years, although the 
number of fish observ^ed may be small. 
The oldest hatchery in the State of New York is at Caledonia. The water supply 
of this hatchery comes from springs located about a mile distant from the hatchery. 
The water supply is unusually ample and there are maintained at this hatchery about 
30,000 adult fish. Each year at spawning time approximately 25 to 50 visible tumors 
are found in handling the fish. This condition has obtained for about 25 years, being 
within the memory of the present superintendent. The fish are 10,000 rainbow trout 
and 20,000 brown trout. The brown trout are the offspring of the first importation of 
