256 
BULLETIN OF THE BUREAU OF FISHERIES 
In the samples taken in each layer we find a total of 192 spat in the top, 138 in the 
middle, and 174 in the bottom layer which shows fairly even distribution in each zone 
as a whole. The spat were not more numerous in the bottom layer in the crates 
because the close fitting of the lath on the bottom and lower side apparently inter- 
fered with the penetration and setting of the larvae. However, if we take the shells 
in the corners or on the outside edges of the crate we find the usual decrease in setting 
above low-water mark from 69, at the bottom of the crate, to 57, at the middle or 
halfway up, and 50 at the top. 
A study of the effect of close grouping of crates on the distribution of spat within 
them was carried out in Group III where the crates were set out within a few inches 
of each other. (Fig. 34.) The crates on the outside edges of this group collected an 
average of 1,600 spat per bushel of shells while those in the middle were even more 
efficient and contained from 2,500 to 3,500 spat on the same quantity of material. 
An increase in the intensity of setting in the middle portion of the group can be attrib- 
uted to the decrease in the velocity of the tidal currents and the creation of eddies. 
In the crates as a whole, setting was invariably found to be most intensive on the lee 
side of the collectors where the larvae apparently were able to attach with greater ease. 
The setting in the crates placed more than a foot below low-water fnark (Group VI) 
was extremely light as the shells planted there became foul and covered with silt in a 
very short time while those which are above low-water mark are kept clean by regular 
exposure to the sun and air. With the conditions existing in Milford Harbor the best 
results can be obtained by placing spat collectors on areas that lie in a zone extending 
from 1 foot below to 2)4 feet above low-water mark. Such areas are rather limited in 
northern waters but by using the lath crates or a similar constructed device, sup- 
ported by legs sufficiently long to hold them above the bottom, it is possible to use 
the same area for the planting of spawners and for setting out collectors. 
EXPERIMENTS IN 1927 
In continuation of the work of finding a cheap, efficient, and practical spat col- 
lector, wire bags filled with shells, as shown in Figures 2 and 37, were tested out in 
1927 in Milford Harbor and Great South Bay, Long Island. Six hundred collectors 
of this type were planted on the flats in Milford Harbor and in addition 300 bushels of 
shells were put out in a pyramidal wire collector having a wood base 12 inches square. 
This latter type of container was designed by Capt. Charles E. Wheeler of the Con- 
necticut Oyster Farms Co., and proved to be suitable for the planting of shells on soft 
mud bottoms. One thousand bushels of oyster shells were scattered over the bottom 
between the bags in order to make a comparison between this usual method of seed 
collection and the wire bags. 
Because the spring and early summer water temperature in 1927 were below 
normal, the oysters developed only a small quantity of spawn and were not fully 
ripened until near the middle of July. Spawning occurred on July 24 and the setting 
of the larvae reached its peak on August 8. The setting was slightly heavier than in 
1926 averaging 2,450 spat per bushel in the wire bags and 1,500 spat in the pyramidal 
type of collectors. The number of spat collected in the bags was found to vary from 
approximately 1,500 to 3,500 per bushel, according to the location in which they were 
planted in the harbor, or to their position in relation to low-water mark. 
The distribution of spat within the bags was very satisfactory, the setting being 
fairly uniform and occurring on over 95 per cent of the shells. In making the counts 
