61 
of another German investigator, Dr. Droscher, of 
Schwerin, indicate that one of these, the roach, is some- 
times largely an animal feeder. 
It is probable, therefore, that we have no vegetarian 
feeders among the fishes yet subjected to cultivation in 
America, or among those that are of sufficient import- 
ance to call for cultivation in the future, except possibly 
as furnishing food for more valuable species. All the 
salmon and trout and whitefish the grayling and smelts, 
ie perch, the bass; dc., must be regarded as. strictly 
animal feeders. 
Mr. Susta’s classification is liable to this criticism, that 
though it may apply correctly to the adult stages of all 
the fishes mentioned, it does not recognize the fact that 
when very young, all the predatory fishes must perforce 
take their food in very small mouthfuls, and are there- 
fore in their early stages to be classed with carp and 
tench and the coregonoids as subsisting on minute 
animal life. Now the great problem of fish-feeding with 
us is, at present, and perhaps will always continue to be, 
how properly to nourish the fish through their early 
stages; so, whether we are rearing carp or trout or 
salmon, the question of providing minute animals for 
their sustenance is one of paramount intérest. 
While Susta was studying the digestion of his carp, 
another fishculturist, Thomas Dubisch, the manager of 
a large pond-interest in Austrian Silesia, was perfecting 
and reducing to practice a system of management for 
carp-ponds, ‘of his invention, which has attracted. great 
attention, and is likely to be generally adopted by ¢ carp- 
breeders ; and from which we can sefely draw lessons 
for our cuidance in the culture of salmonidae in ponds. 
The Dubisch method, or that portion of it with which 
we have to do, consists essentially in having a soft and 
rich bottom to the pond, and drawing off the water so 
that the pond shall be completely empty and dry as 
long as possible each year, including the winter. If a 
