CLASSIFICATION OF THE CEPIIALOCHORDA. 287 



tion." In 1897 1 Willey, in describing A. caudatum, says, 

 " the specific differences between species of Amphioxus 

 are frequently of apparently little moment, but may be 

 of importance when taken in conjunction with the geo- 

 graphical distribution," and in a foot note he adds, " on 

 this principle the Japanese Amphioxus, recently described 

 by Dr. Andrews, should at least be regarded as a marked 

 variety of, rather than identical with, Amphioxus 

 belcheri;" and he suggests the name A. belcheri, var. 

 japonicum. 



In 1897 Nakagawa 2 published most detailed notes on 

 the variation in this lancelet, on a study of over 

 a hundred specimens, but he hesitates to call it B. belcheri 

 because he had not been able to compare the two, and so 

 leaves it as Amphioxus sp. 



In 1901 Jordan and Snyder, 3 in reviewing the lancelets 

 of Japan, constitute a new species for this form. The 

 characters given by them as distinctive of this species 

 could, however, be applied with equal truth to B. belcheri. 

 The myotome formula given has also been recorded for 

 B. belcheri. The reason given for forming a new species 

 is the geographical distribution. They go even further 

 than Dr. Willey, for while he suggested that it was a 

 new variety only, Jordan and Snyder consider it a new 

 species solely on account of its geographical distribution. 

 They say, " In view, however, of the almost entire differ- 

 ence in species .between the shore fauna of Japan and 

 that of Borneo, it seems to us best to regard the Japanese 

 lancelet as a species distinct from B. belcheri. It needs 

 comparison with no other." It will be seen in the first 

 place that this statement distinctly contradicts that of 

 Dr. Andrews, as to the similarity of the shore faunas of 



1 Q.J. M.S. Vol. xxxix. 2 Annot. Zool. Jap. Vol. i., part 4. 



3Proc. U.S. Nat. Mus. No. 1,233. Vol. xxiii. 



