LARVAL OR SWIMMING STAGES 35 



the depths filled in and sketches made with a drawing apparatus. After 

 a good number of these had been made, a comparison both of the out- 

 lines and of the measurements would indicate whether a given size fell 

 naturally between the preceding and succeeding ones, and consequently 

 whether all members of the series seemed likely to belong to one species. 

 Wherever it was necessary the greatest care was taken in reading measure- 

 ments, even to the splitting of the micrometer lines and in using the same 

 longitudinal and vertical axes, the larvae being placed in the position of a 

 creeping clam with the umbos above and the foot below. With ocular V 

 and objective 4, the measurements for the shell of a free-swimming 

 larval mussel, for example, extend all the way from 15 to 58, each unit of 

 which as determined by a stage micrometer representing a value of 6 . 9 m- 

 A pretty complete series of this species was formed, which agreed with 

 drawings and measurements I had made of young mussels found attached 

 in the axils of rock-weed (Fucus) at St. Andrews, when I was studying 

 the clam-fishery in 1900. 



Having determined the mussel, it served as an excellent guide in 

 pursuit of others, and at the same time simplified matters by eliminating 

 from the field of research the most overwhelmingly abundant of all bivalve 

 larvee. It was found that the distribution of plankton stages of the 

 mussel corresponded with the distribution of adult mussels, and this 

 suggested a f aunistic study of bivalves as a means of determining the most 

 probable larvae to be expected. At Malpeque, Mytilus, Mya, Ostrea and 

 Venus are all plentiful, while Clidiophora, Anomia, Mactra, Modiola, 

 Pecten, Saxicava, Macoma, Ensis, Yoldia, Cardium, Tottenia, Kellia 

 and others occur in considerable numbers. It was apparent that there 

 were many species of bivalve larvae in the plankton, but there was nothing 

 in them to directly betray their affinity with the adults, and I had no 

 means of referring them with precision to their proper species. Under 

 such conditions no certain steps of progress could at the time be made by 

 following up the comparison, so I turned to experiment. I set out below 

 low-water mark crocks containing strips of glass held separate by wire 

 networks and in a short time had the satisfaction of finding, fixed to the 

 glass, minute spat stages, in the umbonal regions of which could be recog- 

 nized the familiar plankton shell I had already suspected to be that of the 

 oyster. New observations came thick and fast. To follow the free- 

 swimming plankton larva of the oyster into the fixed stages of the spat 

 was not enough. It became desirable to discover also what its different 

 companions were, for, by proving that they were the young of other 

 species of bivalves, it would make the case of the oyster more secure. The 

 larva of Anomia was determined in a similar way to that of the oyster. 

 The larva of the clam was kept under surveillance during the following 

 seasons at Gaspe, but was not definitely determined until two years later, 

 when I returned to St. Andrews, the great centre of the clam fishery. 



