REPORT ON THE CEPHALOPODA. 75 



alternately distending and contracting with the respiratory movements, will easily under- 

 stand how variations in form may arise. Nevertheless, when a number of specimens of 

 the same species are compared which have been subjected to pretty much the same 

 treatment, a certain similarity is usually to be remarked among them. 



The relative length of the arms, both with respect to each other and to the body, was 

 regarded by d'Orbigny as a point of the greatest importance, and although it has again 

 and again been shown that he greatly exaggerated this,^ nevertheless he has been followed 

 by many authorities, such as Gray and Tryon. The arms of an Octopus being almost 

 wholly muscular, and entirely devoid of any hard or even cartilaginous parts whatsoever, 

 are capable of a very considerable degree of contraction and extension, and as, when 

 killed by being placed in fresh water or alcohol or poisonous solutions, they twist their 

 arms about in the most lively manner, it follows that they may die Avith them in very 

 varied states of contraction. There are, of course, cases in which the dimensions of one 

 or more pairs of arms are so preponderant that no hypothesis of irregular contraction will 

 account for the difference, which must then be regarded as a matter of systematic import- 

 ance ; such instances are to be seen in Octopus macropus, Risso, where the first pair, 

 and in Octopus aranea, d'Orbigny, where the fourth pair of arms greatly exceed the 

 others. When, however, d'Orbigny jDlaces his Octopus fontanianus in the division 

 "Bras inferieurs les plus longs," because the length of the upper arms is 165 mm. and of 

 the lower 166 mm., he carries out the principle to an extent which is almost too ridiculous 

 to require criticism. 



The degree to which the arms are united by a web or umbrella is a valuable 

 character, though it must be borne in mind that here, as in the case of the arms, small 

 variations must be regarded as probably due to different degrees of contraction. 



The colour has commonly been regarded as of but little systematic value, owing to 

 the manifold variations in this respect which these animals undergo owing to the play 

 of the chromatophores. It seems, however, reasonable to suppose that in animals which 

 have been killed and preserved in the same way, the chromatophores will be similarly 

 aflfected, and thus the differences which are due to their action eliminated. Certainly in 

 examining the Challenger material I have remarked that specimens which seemed on 

 other grounds to be referable to the same species have generally agreed also in regard to 

 colour. There are some cases {e.g.. Octopus pictus, Brock, and Octopus lunulatus, Quoy 

 and Gaimard) in which the colour is disposed in definite bands or patches, and in these 

 its distribution is cjuite diagnostic. 



The nature of the surface of the body, namely, whether smooth or provided with 

 warts or cirri, is also regarded by d'Orbigny with disfavour, on the ground that variations 

 in this respect occur in correlation with the position and state of irritation or repose of 

 the animal. The remarks just made regarding the colour apply to a large extent here 



1 See Verrill, Ceph. N. E. Amer., p. 381, and pp. 86, 93, UX) of the present Report. 



