110 Papers from the Marine Biological Laboratory at Tortugas. 
breeding-swarm, and if the lack of proper circulation is still greater the 
worms may still live, but their sexual products atrophy, and they do not 
react to the stimulus which calls forth the breeding-swarm in worms living 
under normal conditions. 
Worms living in rocks which are placed in a floating live-car will still 
swarm, even though they be in a “tideless sea.” Thus rocks containing 
full-grown worms were placed in a scow-shaped live-car 2 meters long, 1 
meter wide, and 1 meter deep. This scow was allowed to float half full of 
water upon the sea, being buoyed up by barrels. Ample circulation was 
secured by holes in the sides and bottom, and a white canvas awning served 
to shield off the rays of the intense noon-day sun of the tropics. This scow 
was entirely open, so that moonlight fell freely upon it. In 1905 and in 
1908, the rocks containing the worms were allowed to remain in the floating 
live-car for 30 days previous to the date of the normal swarm, but in 1907 
they were placed in the floating scow only three days before the normal 
day of swarming; but in all three experiments some of the worms swarmed 
normally within three days of the day of the last quarter of the July moon. 
Altogether, 4 out of 11 mature worms swarmed normally in these tideless 
live-cars, but 7 of them did not cast off their posterior ends, but remained 
passive in their burrows. 
In nature all of the mature worms swarm at the annual breeding-time, 
and this partial failure of the worms to swarm may indicate that the chang- 
ing pressure due to rise and fall of tide over the reefs is a contributory but 
not a@ necessary component of the stimulus which calls forth the breeding- 
swarm. It is more probable, however, that confinement within the wood- 
inclosed space of the live-car and the lack of perfect circulation of water 
acted as a partial preventive of the swarming, and that the reaction is wholly 
independent of the rise and fall of the tide. In any event, it is evident that 
the worms can swarm normally in a tideless sea, and that rise and fall of 
tide is not a necessary or sole cause of the swarming. 
On the other hand, the worms have never swarmed when moonlight was 
prevented from falling upon the rocks within which they lived. In order to 
test this, I had floating scows similar to those used in the previously de- 
' scribed experiment, but they were provided with light-tight wooden covers, 
so that they could be closed at sunset every evening and opened soon after 
sunrise every morning, thus preventing the moonlight from falling upon 
the rocks. Altogether I had at least 22 mature worms in the rocks within 
these darkened live-cars, and in 1907 the moonlight was excluded for 5 days, 
in 1906 for 14 days, in 1905 for 30 days, and in 1908 for 2 days before 
the date of the swarm; but none of these worms showed any indication 
of swarming, and it appears that they could not respond, owing to the 
absence of light. It is probable, therefore, that the worms can not swarm 
unless moonlight falls upon the rocks. In nature the worms will swarm in 
