Vol. 50. J SYSTEMATIC POSITION OP THE TRILOBITES. 429 



verse and longitudinal sections of Apus, I am convinced that the 

 basal regions of the limbs of trilobites were membranous lobes with 

 long transverse insertions, which probably passed laterally into the 

 membranous under -surfaces of the pleurae. Fig. 15 is, I think, 

 completely explained by this supposition. 



These membranous basal plates probably projected inwards 

 towards the ventral median line all along the trunk, as they still do 

 in Apus, perhaps as segmental repetitions of the masticatory plates 

 round the mouth. In Apus, I believe, they are still functional, 

 and serve to push food forward towards the head and mouth. The 

 anterior pairs are armed with teeth, and foreshadow the maxilli- 

 pedes of the higher malacostraca. 



We conclude, then, that the limbs of the trilobites, in spite of 

 their development of filiform ambulatory legs, were originally mem- 

 branous lobes, and that their basal regions persisted as such. This 

 is of primary importance, as it places their affinity with the phyl- 

 lopods beyond question. 1 



Of equal importance is the fact which I have elsewhere already 

 insisted upon, that the limbs of the trilobites show the same 

 gradual diminution in size from front to back which we find in 

 Apus, the most posterior being quite minute and rudimentary. If 

 my explanation of this remarkable phenomenon be correct, namely : 

 that these posterior segments are fixed in an undeveloped larval 

 condition, then these early phyllopods were clearly not very far 

 removed from ancestors with a very much richer segmentation than 

 they themselves possess, or than Apus possesses. Apus cancriformis 

 develops, or commences to develop, upwards of sixty segments, and 

 may thus well be descended from a form with seventy to eighty, 

 or even a hundred segments. 



Summary. 



It is now possible, from the foregoing considerations, to fix with 

 great probability the zoological position of the trilobites. The 

 bending round ventrally of the first segment, the great labrum with 

 antennae attached at its sides, the ' wandering ' of the eyes, the pores 

 (pointing to the probable presence of water-sacs), the head with a 

 varying and progressively increasing number of segments, the dorsal 

 organ, the rudimentary character of the posterior segments, and the 

 gradual diminution in size, with the essentially lobate or phyllopodan 

 type, of the limbs, all serve to connect the trilobites with Apus. 



This relationship cannot, however, be considered as direct. Apus, 

 on account of its richer segmentation, the absence of pleurae on the 



1 [Since this paper was read, Dr. Eeecher has described the ' Appendages of 

 the Pygidiurn of Triarthrus,' Amer. Journ. Sci. ser. 3, vol. xlvii. p. 298, April, 

 1894. The limbs of the rudimentary pygidial segments of Triarthrus are almost 

 indistinguishable from the rudimentary limbs of the larval segments in a growing 

 Apus, which till now were unique among the limbs of arthropods. The limb3 of 

 trilobites, whatever their adult form, were therefore beyond question develop- 

 ments of originally phyllopodan appendages. Their transitions from front to 

 back, that is, from filamentous to membranous, is also exactly paralleled in 

 Apus, see figs. 9, 4, 5, and 10 in 'The Apodidae.'— H. M. B., June, 1894.] 



Q. J. G. S. No. 199. 2 a 



