THE OSTEACODA. 



279 



Archiconchcecia ctjcullata, Brady. 

 Judged by a measurement of liis fig'ure, Dr. Miillcr appears here, as elsewhere, to 

 have inchided the rostrum and posterior spine in his measurements : — ? 1-85-2-2, 

 e 1-8-2-2 mm. The measurable male of my two specimens (32 I) was only I'S mm. 

 by my method, 1-9 with the rostrum, 2-1 with the rostrum and spine; it fell, therefore, 

 into the same stage as Miiller's specimens. The two females measured I'O mm. by 

 my method (1-4 with the rostrum and spine), and belonged therefore to an earlier 

 stage, II. or III. 



CYPRIDINID^. 



Oypridina castanea, Brady. 

 Apparently three stages, I., II., and IV., were represented in the five specimens ; 

 with a growth-factor of 1-5, the lengths 1-8, 27 (missing), 4-0, and 6-0 are related as in 

 the other species. 



Ser7'ich's Records o/Homarus americanus. 



It will, I hope, be apparent to the reader that the species just enumerated do respond, 

 however imperfectly, to Brooks's law. In Herrick's great work on the American 

 Lobster (Bull. U.S. Fish Commission for 1895) are given records which enable us to 

 test it in another group of Crustacea. For this purpose his " Becord of moults of larval 

 and adolescent stages " (table 31-, pp. 176-177) is the most significant. The first three 

 columns in the following table are taken from this record ; the fourth column shows 

 the product of the mean length at each moult X a growth-factor 1-14, which is in every 

 case remarkably close to the next highest observed mean length. 



It is quite evident that the same relation exists in tlie early stages of Ilomarus 

 americanus as in the Halocypridai. The actual ratio is much low er in the lobster than 



