562 



Phylum Chordata 



Fort Jefferson. Having been torn from their attachment, individuals 

 soon die but may be kept in a live-car for a day or two without deteriora- 

 tion of their eggs and sperm. Fertilization of the eggs and development 

 of the relatively small larvae take place outside the body of the parent. 

 This larva has an average total length of about 0.93 mm. (body length 

 0.18 mm., tail length 0.75 mm.). Development is exceedingly rapid. 

 At temperatures between 27 and 29° C, eggs fertilized at 5 a.m. 

 yield free-swimming larvae at about 1 p.m., and one lot hatched after 

 a period of development of only 6 hours and 38 minutes. By fertilizing 

 eggs taken from several individuals early in the morning it is possible 

 to have many thousands of free-swimming larvae available for experi- 

 mentation during the afternoon. 



The procedure followed with success is simple. After cutting away 

 the projecting ends of the siphons, incisions are made in the tunic be- 

 tween the siphons and along one side from siphon to base of body. 

 Spreading the opening thus made and seizing the stump of one siphon 

 with forceps, the body is pulled away with no injury save to small blood- 

 vessels. Removed from the tunic, washed with fresh seawater and 

 placed separately in dry petri dishes, animals were left until the blood 

 from severed mantle vessels had clotted. Since the long genital ducts lie 

 close together near the surface at one side of the body and, when filled 

 with eggs or sperm, are plainly visible, it was then possible to puncture 

 the oviduct near its opening into the atrial cavity without rupturing 

 the sperm duct, provided the puncture was made with a very sharp 

 needle. As eggs flowed from the puncture they were taken up with a 

 wide-mouthed pipette and transferred to beakers of fresh seawater and 

 later fertilized with sperm taken either from the same or a different indi- 

 vidual. Especial care had always to be taken that no blood was taken 

 up with the eggs or sperm as even slight contamination results in ab- 

 normal development. 



For experimental studies in which the statistical method of measuring 

 results must be applied, it is desirable that the factors of individual 

 variation be reduced to a minimum. This desideratum is approached 

 in larvae of ascidians when each experimental lot is composed of indi- 

 viduals having the same parentage; even self-fertilization adds to the 

 desired uniformity. But eggs of different parentage develop at rates 

 sufficiently different to permit one lot to hatch several minutes before 

 or after another, and undoubtedly they vary more or less in all hereditary 

 characters. 



It was often necessary to divide eggs taken from a single individual 

 and to place them in two or more beakers, for when too many eggs were 

 crowded together during development it was found that abnormalities 

 appeared during the later stages, often to the extent of 100%, depending 



