Mr. L. J. Cole on Tin-leggpd Pi/ennaonids. 409 



knowledge of tlie embryology is going to beof more importance 

 in elncidating the question of the relationship of the 

 Pycnogonida to the other Arthropoda. 



Tn the following discussion 1 have used the families as 

 adopted by Sars (iHgi), not because I consider them neces- 

 sarily more true to a correct system than those given by 

 some other authors, but because they are better adapted to 

 my present purpose. 1 shall use them simply as milestones 

 to mark certain points, or, perhaps better, stretches along 

 the road of phylogeuetic development, without intending to 

 imply that these stretches might not consistently be fewer or 

 greater in number, and with their limits at different points. 



The systems given by most recent authors differ from that 

 which 1 have temporarily adopted here for the most part 

 only in the employment of different names^ derived from 

 other genera, or in being more or less inclusive. The number 

 of families given by Sars (1891, p. 144) is larger than is 

 generally recognized, and conseqviently his classification is 

 more useful to me in showing the gradual stages of differen- 

 tiation ; and where, in one or two instances, this series has 

 not been complete enough for my purpose, I have introduced 

 generic names for greater completeness. In most cases the 

 names of the families ai'e derived from well-known genera; 

 where another genus in the family is perhaps better known 

 I have indicated the fact. I have retained the order of 

 sequence as given by Sars, except fur the transposition of the 

 Pasithoidse to a position next the Ammotheidae instead of 

 the Eurycydidfe, as they appear to be more nearly related to 

 the former. 



The character on ^hich the families are chiefly based is 

 the presence or absence or the state of development attained 

 by the first three pairs of appendages — chelifori, palpi, and 

 ovigera. In general those forms which possess these ap- 

 pendages well develojjed are to be considered the more 

 primitive, a fact shown by the ontogenetic development of 

 those which do not possess them in the fully adult stage. 

 The condition of these appendages (together with the 

 presence or absence of denticulate spines on the ovigera, the 

 degree of trunk-segmentation, and the number and position 

 of the genital openings) is given for the various families in 

 the table on p. 40>5. 



It will be noticed from the table that there are two 

 diverging series, each starting from a primitive condition, 

 fairly w^ell represented, seeminjily, by Dtcolopoda. Decolo- 

 poda might well be the basis for a distinct family, the 

 Decolopoclidae, if it were not for difficulties encountered in 

 Ann. ^ May. N. Hist. Ser. 7. IV. xv. 28 



