Howard. — Glochidia Infection on River Herring 97 



unusually wide lamellae on the gill-filaments. The 

 glochidium in clamping upon these, will include usually 

 four or more lamellae within the bite of its valves. Thus 

 several lamellae will be enclosed in a cyst which is for 

 the most part a solid mass of cells. Sometimes the space 

 between lamellae persists proximally within the other- 

 wise homogeneous and solid cyst. Such spaces would ex- 

 plain the phenomena which Surber (Surber, 1913, p. 

 Ill) called ''migration of the glochidium," and further 

 described as follows, "Simple, hollow, globular areas, sug- 

 gesting small tumors, unmistakably the former abode of 

 some young mussel." The same globular areas are to be 

 seen in the herring, in infections by glochidia of the Ele- 

 phant-ear mussel, Elliptio crassidens (Ort.). The more 

 simple explanation which we have offered seems to be 

 supported by the appearances in many cases and to make 

 necessary the theory of migration which is unknown for 

 larval mussels on other fish. 



Eight days after the last shipment of gills from the 

 impounded fish, a report was received that the remaining 

 Pomolobus had escaped. This unfortunate occurrence 

 prevented the determination of the full period of para- 

 sitism. The writer has determined that this period is 

 the same for three different species of mussel, (Lampsilis 

 ligamentina, L. luteola, and L. anodontoides) under like 

 conditions. (Howard, 1914b). The period was 20-21 

 days (May 22 to June 10 and 11) — that is, the three 

 were carried parallel through the metamorphosis, having 

 the same date of infection and being held in the same 

 pond. To that extent, conditions were similar. The one 

 factor varied was the use of different hosts. This sum- 

 mer I observed a short period of 7 days for Quadrula 

 plicata, but at the average high temperature of 90° F. 

 (32° C). I have seen indications of a difference in the 

 parasitic period for widely separated species but know 

 of no demonstration of it. The amount of development 

 during the ten to twelve days of observations in the pres- 

 ent experiment, indicates that the period of metamorpho- 

 sis in Quadrula ebenns corresponds rather closely with 



