308 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



Thus in one segment of the 9-radial example there are 16 interradial 

 tentacles; in another segment only 6. In the 7-radial specimen 

 there are 9 tentacles in one segment ; 11 large and 4 small in a second ; 

 and 7 large and 3 small in a third. 



There are one or two otocysts between every two tentacles ; that is, 

 many more otocysts than tentacles. I was not able to count the 

 entire number in either specimen, because part of the margin was 

 damaged in each. 



The gonads are spindle-shaped, thickest distally, and are re- 

 stricted to the outer one-half or one-third of the radial canals. They 

 do not quite reach to the circular canal. 



Color. — In one specimen the gonads are brownish-yellow; in the 

 other pale pinkish. There are no notes available as to their color in 

 life. 



Comparison between these examples and Maas's Amboina speci- 

 mens shows that they agree even to minor details. 



Family AEQUORIDAE Eschscholtz, 1829. 



LEPTOMEDTJSAE with very numerous radial canals and with closed otocysts. 



These are the family limits which have been adopted by Maas, by 

 Browne, and the writer, for the presence of more than eight canals 

 can not be used as the dividing line between eucopid and aequorid, 

 as I formerly proposed, because occasional specimens of eucopids 

 may have as many as 11 (p. 307, Browne, 1905). 



Mayer (1910) extends the family to all Leptomedusae with eight 

 canals, irrespective of whether the otocysts are open or closed; that 

 is, he includes Octocanna, Octogonade, and Halopsis. But the for- 

 mer (p. 306, Maas, 1905) certainly belongs to the Eucopidae; indeed, it 

 is a question whether its " species " are anything but variant tetra- 

 nemal eucopids. Octogonade and Halopsis have open sense pits, 

 with ocelli as well as otoliths, hence belong to the Mitrocomidae. 

 Browne (1910), it is true, says that reinvestigation of the sense 

 organs of Halopsis is necessary to show whether they are really open. 

 But I have recently been able to establish, in fresh material, that 

 they are open (1914, p. 102). 



No family corresponding to the Mitrocomidae of Torrey (1909) 

 and Browne (1910) is recognized by Mayer, the genera listed above 

 being distributed between Eucopidae and Aequoridae, according to 

 the number of their canals. But the structure of the sense organs 

 shows that the Mitrocomidae are a natural group. 



The Aequoridae are still perhaps the most puzzling family of 

 Leptomedusae to the systematist; no division into genera which has 

 yet been proposed covers the ground adequately, nor are the normal 

 limits of variation yet known for a single species. Yet the family 



