MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 519 



From the few records yet available, it is impossible to determine how far the different 

 species of corals have any regular breeding season; such would scarcely be expected under the 

 uniform conditions of temperature characteristic of tropical waters. During the month of Julj' 

 four or live different colonies of Siderastrxa radians were collected, all having polyps charged 

 with free larvte. while other collections made at different times from the same locality never 

 yielded fertile colonies. From another locality larvse of this species were secured during the 

 middle of March. 



Most of the larvte to be here described were collected in the early months of the year (March, 

 April); but larvie of Manicina areolata and Favia fragum seem to be extruded nearly the whole 

 year round. 



SPERMARIA AND OVA. 



In their detailed structure, the spermaria and ova of corals are much like the corresponding 

 elements in the Actiniaria. A mesentei\y bearing two spermaria, in addition to three ova. is 

 represented on PI. XX, tig. 140, taken from Mxandi'ina. The fertile portion of the mesentei\v 

 is greatly swollen, and the mesogliea is thin and surrounds each ovum and spermarium as a verj' 

 delicate sheath. The mesenterial epithelium has undergone certain moditications: toward their 

 base the cells are much vacuolated, and without any recognizalde protoplasmic contents. Cell 

 limitations are not obvious, and the nuclei are arranged in a very deffnite zone toward the 

 margin, along with the other protoplasmic contents of the cells. 



The spermaria are crowded with deeply-staining, spheroidal, sperm cells, each with several 

 still more deeplj'-staining particles or nuclei. A similar stage has ])een figured by Professor 

 Hickson for Alcyonium (1895), and also described by Ashworth (1899) for Xenia. Occasionally 

 a central cavity containing a coagulum is already developed, and here and there this is tilled with 

 the projecting tails of the spermatozoa. Without any doubt the ova and spermaria are developed 

 from the mesenterial epithelium, as in other Anthozoa, and the cells wander into the mesoglcea 

 and become encapsuled by it. but none of the earliest developmental stages have been found. 



The spermaria fi'om another colony of 3/xaiuhi»a, instead of being nearly spherical, are 

 narrow, and extend along the transverse length of the mesentery for some distance, and display 

 constrictions, as if made up of four or five fused spermaria. 



On one occasion, ripe spermatozoa were observed in the act of extrusion from a polyp of 

 Favia fragum, the process taking place by the same jerking motion as when larvfe are liberated. 

 They were of the same form as those of Actinia equina, described and figured by Lacaze-Duthiers. 

 that is, a pear-shaped head with a laterally fixed tail. 



A mesentery bearing two large ova is represented in fig. I-IO, again taken from Msemidrina. 

 The eggs contain a large amount of vacuolated, finely granular yolk. The vacuolization is very 

 uniform except peripherally, where the ovum stains more deeply, owing to the greater concen- 

 tration of the protoplasm. In the same region deeph'-staining granules are also numerous. The 

 germinal vesicle is large and homogeneous in structure, and is unaffected hy carmine stains, but 

 readily takes up aniline blue. The germinal spot is usually situated close to the margin of the 

 germinal vesicle, and differs from the latter in staining intensely in liorax carmine. 



AVhite, spheroidal, unfertilized eggs were extruded singly from Facia fragum on sevei'al 

 occasions, after floating around in the gastric cavity for several hours. Examined under the 

 microscope, the ova underwent a great variety of irregular movements for about an hour, sending 

 out lobate processes, tirst from one region and then rapidly from another. 



The experiences of von Koch, Wilson, Haddon, and those here recorded render it manifest 

 that the eggs in coral polyps are ripened in batches, not a few at a time, and that the larva» 

 develop equally; for in all cases where larvaj have been obtained from any polyp thej' occurred 

 in numbers, and practically of the same age, while many far advanced eggs were present in 

 the mesenteries preparing for another series. This is usually the case in the Actiniaria, l)ut 

 excej^tions may occur, as where larviv at all stages have been .secured from the same polyp at one 

 time. Fowler (1888, p. 13) states that the ova in Sj^henotrochus ruhe^cen)s were in various stages 

 of maturation. 



