Report on the Hydroids. 343 



Tiergruppen kann man lur solche Untersuchungen leichter Zählungen 

 oder Masse verwenden, für die Hydroiden aher muss man die viel 

 mehr zeitrauhenden Projektionsmethoden meistens benutzen", (1909 

 (17), p. 135), I do not think he is right. We might be able to make 

 measurements in such a manner, that we could obtain numerical 

 expression for the special characteristics of the species, 

 and if the material is sufficiently large such measurements would 

 be an important help to the limitation of the species. 



The projection method leaves too much freedom for a conscious 

 or unconscious arbitrariness in the choice of the variants to be 

 drawn, and can never give very trustworthy pictures, either of the 

 extremes or of the norm for the species; the last would be of import- 

 ance especially in biological regards in that it shows, how the species 

 is typically developed in this or that locality under this or that set 

 of natural conditions. A comparison of material from different places 

 can only lead to a trustworthy result, when we have numbers to 

 rely upon. An ingenious investigator may perhaps be able to draw 

 far-reaching conclusions by simply glancing at a series of drawings; 

 but his conclusions cannot be controlled by the less ingenious, who 

 may desire more solid grounds for accepting the result as trust- 

 worthy. I hope in time to have the opportunity of going further 

 into this matter and of using the numerical method in practice. In 

 the present paper I have made a small endeavour in this direction 

 but unfortunately the material is not very large. 



To Inspector G. M. R. Levinsen, who entrusted this work to me 

 and has helped me with advice and instruction, I would offer here 

 my most sincere and hearty thanks. 



