244 
BULLETIN OE THE UNITED STATES FISH COMMISSION. 
(Miilleria), so common in parts of the Bahama Islands. It was with the greatest 
difficulty that a few embryos of this form could be got in the laboratory. When, how- 
ever, the animals were confined in a large box anchored in about a fathom of water, 
quantities of developing eggs could be had by drawing up with a tube some of the 
sediment in the bottom of the box. 
The “spat” once obtained in abundance, success will next depend largely on the 
selection of the locality in which the young sponges are to be set out. A careful study 
of the Florida grounds should be undertaken, with the view of investigating, among 
other points, this very matter of the kinds of locality best adapted to the growth 
of the various grades of sponges. Quiet water, a firm bottom, and an absence of 
muddy sediment seem essential desiderata. The question of enemies is probably 
of minor importance, and yet the well known student of sponges, Vosmaer, mentions 
that he has several times seen the European hermit-crab ( Pagurus ) greedily eat a 
common silicious sponge ( Suberites ), certainly quite as unappetizing a morsel as 
the coarsest commercial sponge. 
When it has once been accurately determined what are the physicial and biolog- 
ical characteristics of the Florida grounds, which produce the finest sponges— and it 
may be mentioned here that sponges are among the most variable of animals and 
seem to be peculiarly affected by their surroundings — a detailed comparison should 
be made between these grounds and those parts of the Mediterranean producing the 
finest grades. The purpose of such a comparison would be to discover whether we 
rerally lack any of the natural advantages necessary for the production of the finest 
sponges and, if so, whether these can be artificially reproduced — whether, for instance, 
it would be possible or desirable to imitate on this side a particular kind of bottom 
found in the Mediterranean. 
Following on the investigation of the sponge- grounds, I believe it to be eminently 
desirable to start a series of experiments, the purpose of which shall be to discover 
how far, along what lines, and by what means sponges may be artificially altered by 
breeding. The great variability of sponges in nature leads one to believe that they 
would quickly respond as individuals to a change in the environment, and thus, simply 
by growing the animals in a superior locality, an improved variety, constant, as long 
as the sponges continue to grow in that locality, might be produced. It is quite 
likely that such improvements could be carried out on sponges propagated by cuttings 
as well as on those grown from eggs. In improving races, however, it has always 
been found that the two important means are sexual breeding from selected specimens 
and grafting, the latter method being commonly regarded as only applicable to plants. 
In sponges, as in other organisms, increase of knowledge will in all probability 
confirm the belief', already fairly well grounded, that individuals developed from the 
fertilized eggs vary more, i. e., exhibit more differences one from the other, than indi- 
viduals grown from buds or cuttings. Herein, to my mind, lies the advisability of 
growing sponges from eggs as well as from cuttings. The latter method, being quick, 
sure, and simple, can at once be made of great practical use. Breeding from the egg 
is more complex, and must be carefully tried by competent experimenters. In the 
end, however, I believe that it will lead to great improvements in the quality of our 
sponges. 
I would suggest that, after selection of a proper locality, a small plantation of 
sponges developed from eggs be started and carefully watched. As the sponges grow, 
it would be a simple matter to pick out those individuals in which the fiber varied in 
