Ortmann: Families and Genera oi Najades. 311 



narrower than the normal water-tubes, but have much heavier septa, 



with (he usual marsupial structure of the epithelium. When charged 

 the ovisacs swell so as to be considerably wider than the normal 

 water-tubes. The marsupium protrudes beyond the original edge of 

 the gill to a considerable degree, and the single ovisacs are subcylin- 

 drical, and are somewhat curved backward. The ova and glochidia 

 fill the ovisacs in the shape of rather solid placentae, and are packed 

 close together through all of the placental mass (see Lefevre and Curtis, 

 1910, pi. 4, fig. 28). Probably the subcylindrical placentae are dis- 

 charged whole, although this has not been observed. The placentae 

 can be taken out whole (see ibid., pi. 4, fig. 26), and in one of my 

 specimens the holes through which placentae had been recently dis- 

 charged, have been seen (see Plat'' XYIII, fig. 10). The marsupium 

 begins near the base of the gill, so that a considerable part of it is 

 enclosed within the two original laminae of the gill. 



Lefevre and Curtis (1910, p. 97, fig. M) have figured the glochidium, 

 and give its dimension as 0.225 X 0.23 mm. I find that this is sub- 

 stantially correct. The glochidia are of medium size, almost sub- 

 circular (their shape may be best compared with a circle a small 

 section of which is cut off). Length and height about the same: 

 0.22 mm. (see Plate XX, fig. 1). 



Color of soft parts whitish, with the edge of the mantle brownish, 

 chiefly so in the region of the branchial and anal openings. Mar- 

 supium white. 



There is much uncertainty as to the breeding season of this species. 

 Lea (Obs., Ill, 1842) mentions ova as formed in the ovarium in autumn, 

 and Sterki (1898, p. 20} found them in October. Lefevre and Curtis 

 (1910, p. 89) place this species among the forms with a short breeding 

 season, but without giving particulars. The sterile females collected 

 by myself were all found in the month of September, at a time when 

 most other bradytictic forms are gravid. The gravid females from 

 Louisiana, collected by Frierson on August 6, were in part discharging, 

 so that this would tend to show that the breeding season ends at that 

 time. The statements made by Lea and Sterki might suggest that 

 the breeding season begins rather late, in winter, and this assumption 

 would agree with the facts at hand. Yet recorded observations are 

 entirely too few, and attention should be directed to this question. 

 The sterile specimen from Arkansas, collected in February, does not 

 ccntribute to the solution of the question. 



