546 p. W. GAMBLIi AND J. H. ASHWORTH. 



agreement is indicated in Cosmovici's work (1880), in which 

 he describes a gonad attached to the nephrostome, but in 

 such a confused way as to render his conckisions of little value. 

 From the little we know, however, there seems to be a close 

 resemblance between the Opheliidse and the Arenicolidae in 

 reference to the nephridial and reproductive systems. In 

 addition to this, the digestive system and the segmentation of 

 the coelom are also fundamentally similar in the two cases. 

 Here again, however, as in the Scalibregmidfe, the external 

 features, the vascular system, especially the heart, the gan- 

 glionation of the nerve-cord, and, to judge from Kukenthal's 

 account (1887), the brain, are very different from those of 

 the Arenicolidge. 



With the Maldanidge the case is somewhat better. But 

 even in comparing this family with the Arenicolidee the 

 differences are more obvious than the resemblances : for 

 though the absence of conspicuous cirri, the general form of 

 the parapodia, the reduction of the prostomium and ciliated 

 grooves, the small number of nephridia and gonads, the 

 simple subepidermal brain, the ventral cord with giant-cells 

 and a continuous covering of ganglion cells, — all agree fairly 

 closely with the corresponding structures of Arenicola, yet 

 the habits, the external features, and particularly the absence 

 of gills, of otocysts, and of hearts, separate the two families 

 very decidedly. The case of Branchioinaldane discussed 

 above makes it, however, clear that, apart from the seg- 

 mentation of the coelom and the vascular system, of which 

 practically nothing is known in Maldanids, there is a closer 

 resemblance between this family and the Arenicolidse than 

 exists between the lugworms and any third group.^ 



Benham (1893, p. 49) has suggested "that the closely 

 investing gelatinous tube (of the post-larva of Arenicola 

 marina) seems, when taken in connection with sundry 

 internal arrangements, such as nephridia, septa, etc., to point 

 to an affinity with the Chlorhgemidpe." The nephridia and 



1 Mesnil (1898) has suggested an approximation of tlie Arenicolidse and 

 Maldauidse on tlie ground of the similarities of the setse. 



