130 ME. CYRIL CROSSLAND ON THE [June 16, 



distinctions, the proportions of which are described nvimerically 

 and with great minuteness by some of the best workers. How- 

 ever one may admire the care which has been employed, and 

 however desirable the use of such numerical statements may be, 

 my results showed clearly that they are, from the nature of the 

 case, unpractical. As their employment can only render descrip- 

 tions more cumbrous, and the limits of the species so defined 

 more hazy even than in nature, I forego all minuteness in state- 

 ments of proportion and numbers. Accurate figures, drawn to 

 scale, will provide as full and unmistakable accounts of species as 

 is possible by the medium of ink and papei-, the extent of variation 

 rarely being so great as to render a di-awing thus prepared from 

 one specimen not recognisable at once as a likeness of any other 

 member of the same species. 



The deductions fi-om my tables of measurements of specimens 

 of Mmyhysa mossambiea are shortly as follows : — 



(1) The body shape is generally constant, but the numerical 



position of the widest segment may be very diftei'ent in 

 certain individuals. 



(2) The proportions of pro- and peristomium in se and intei- se 



are roughly constant. 



(3) The length of the unpaired tentacle is variable within 



limits, but the number of the segment to which it reaches 

 when laid along the back is not worth giving in a descrip- 

 tion of the species. 



(4) The gills, as often noted, begin on difierent segments in 



difierent specimens, increase in size and comj)lexity quite 

 irregularly, and vary in the maximum number of the 

 filaments they bear. 



In another examination, all the smaller gills were left 

 out of account, the distance from the anterior end to the 

 point where gills of approximately the full size begin to 

 occur being measured. This varies from 5 to 6 cms. ; the 

 number of segments it contains varies from 64 to 80. 



(5) The number of teeth borne by the dental plates frequently 



varies by one on either side of the mean, i. e. for the larger 

 plates from 4 to 6. 



In Marphysa belli, a species which is characterised by the large 

 size and the concentration of the gills to a few segments, these still 

 vary in the same ways. Two specimens, collected and preserved 

 together *, differed as follows : — 



Though both are 3*5 mm. in breadth the segments in no. 1 are 

 much the shorter. In no. 1 the gills do not meet over the back, 

 though no. 2 agrees with the specific definition in this respect. 

 No. 1 bears 17 pairs of gills on feet 13 to 30 t, and no. 2 21 paii-s 



* I owe my opportunity of examining this species to tlie kindness of Prof. Mcintosh. 

 The specimens were dredged by the ' Porcupine ' in 1870 from 81 fathoms, off Cape 

 Finisterre. 



t In all cases the first setigerous segment is counted as the first, a method of 

 reckoning which is free from any possible ambiguity. 



