BIOLOGICAL SURVEY OF WOODS HOLE AND VICINITY. 189 



When subjected to statistical analysis, our results are in full agreement with those 

 of Herdraan, though we are not convinced of the truth of his interpretation. 



As shown in the table on page "jj, the average number of species per dredge haul, 

 as based upon the 458 regular stations of the Survey, is 37.0, while the average number of 

 genera per dredge haul is 34.3. Thus the average number of species per genus is approxi- 

 mately i.oS,° or the ratio of species to genera is about 13 to 12. This ratio we have 

 thought best to compare, not with that derived from a consideration of the entire array 

 of species and genera for the Woods Hole Region, but with that based upon the total 

 number of species and genera, so far as encountered by us in the dredge. Among 

 these are included 510 determined species, representing 361 genera. The average num- 

 ber of species per genus is thus approximately 1.41, or the ratio of species to genera is 

 as 7 to 5. 



So far, then, we seem to be in complete agreement with Herdman. A quite different 

 explanation from that given by him has, however, suggested itself. It has occurred to 

 us that the same relations would follov/ if some of our genera contained a considerable 

 number of uncommon species. These latter might not be taken with sufficient fre- 

 quency to affect appreciably the average number of species per dredge haul, but they 

 would greatly augment the number of species per genus when the total number of those 

 encountered were taken into account. Now, as a matter of fact-, many of the genera 

 do contain a considerable number of rare species — species which were taken once only, 

 or a very few times — in addition to common ones. On the other hand, it is true that 

 we also meet with certain rare species which are the only representatives in local waters 

 of their respective genera. Thus there would seem to be nothing to show whether the 

 inclusion of these less common species would augment or decrease the average number 

 of species per genus in our fauna. 



One test may be applied, however. We may restrict our computation to the com- 

 moner species, and determine the ratio of species to genera among these. For this pur- 

 pose let us employ those species which were taken at 10 or more of our dredging sta- 

 tions. Among these we find 209 species* representing 17S genera. The average number 

 of species per genus is thus about 1.18, a figure very much smaller than that represent- 

 ing the number of species per genus in the entire array of organisms dredged by us. 

 Indeed it approaches more nearly the figure (1.08) expressing the average ratio within 

 the limits of a single dredge haul. Thus the conclusion seems warranted that the 

 larger number of species per genus found to occur in the fauna at large, as com- 

 pared with the average number for a single dredge haul, is due largely, if not wholly, to 

 the inclusion in the former reckoning of the rarer members of certain genera. 

 Such species do not, on the other hand, occur with sufficient frequencv to appreciably 

 affect the ratio for the average dredge haul. If this reasoning be correct, what seemed 

 to be a fascinating and clear-cut demonstration of a significant principle of distribution 

 falls to the ground. 



» We are quite aware that the ratio between these gross averages has not exactly the same value as the average of the separate 



ratios for the various dredge hauls. In other words. ? lll^^JlJSLz. has not the same value as Z*?-!- -+- ) -^3. \\'here the num- 



3 .? \z y z' 



ber of terms is so great, however, the results derived from the two methods of computation must be sufficiently close for present 

 punx)ses. 



^ Including Protozoa. 



