412 MR. H. M. BEENAED ON THE [Aug. 1 894, 



genealogy of the Crustacea 1 from the pen of Prof. Carl Grobben, of 

 Vienna, whose well-known researches into the anatomy and embryo- 

 logy of the Crustacea lend special weight to his conclusions. Prof. 

 Grobben, after reviewing an immense array of facts and arguments, 

 arrives at the conclusion that all the existing Crustacea can be 

 deduced from an Apus-l'ike ancestral form. 



Since the publication of these conclusions, I have been studying 

 the organization of the trilobites themselves, and I wish here 

 to express my warmest thanks to Dr. Henry Woodward, F.P.S., 

 to Prof. Judd, P.R.S., and to Prof. G. B. Howes, for kindly placing 

 specimens at my disposal for examination, and further to Mr. W. I. 

 Last, Keeper of the Mechanical Department at the South Kensington 

 Museum, for the kindly and invaluable assistance he rendered me 

 in fitting up for me a small sandblast, by means of which I have 

 been endeavouring to ' develop ' the fossils. 



I. The great variability in the number of the segments shown by 

 the trilobites need hardly be again insisted upon as a feature con- 

 necting them with the phyllopods. Of still greater importance is 

 the gradual diminution of the size of the segments posteriorly, 

 which remarkable feature the trilobites share with Apus. I have 

 endeavoured to show (op. jam cit.) that this feature is explicable by 

 assuming that Apus is the ' Protonauplius ' of authors, in which a 

 very large number of segments commence to develop, many of 

 which, however, at the posterior end of the body, remain fixed in a 

 rudimentary condition. This explanation of the morphology of 

 Apus is, it seems to me, evident if we compare the adult with the 

 developing larva. The adult is but the grown, not metamorphosed, 

 larva — grown by the continual development of segments from before 

 backwards, until at a certain stage this process becomes fixed, and 

 we have the adult Apus with a number of fixed rudimentary 

 segments. 2 This fixation of a number of undeveloped segments is 

 visible in many trilobites. 



In the early Olenellus these rudimentary posterior segments are 

 still free (i. e. do not form a pygidium). As a rule, however, they 

 form the plate-like pygidium characteristic of the trilobites. This 

 specialization seems to have set in very early ; for instance, in Micro- 

 discus we find a pygidium apparently consisting of only a few 

 segments differing little in size from those of the trunk, whereas 

 a review of the pygidia of the whole order leaves little doubt that 

 this organ was originally composed of a number of larval segments 

 which diminished gradually in size and development from before 

 backward. 



That animals closely resembling Apus were extant in earliest 



1 Sitzungsber. d. k. Akad. Wissensch. Wien, vol. ci. (1892) pt. i. pp. 237-274. 



2 In a recent systematic paper on the genus Apus (Z. w. Z. 5, pt. 1), 

 Dr. Braem records a remarkable inconstancy in the number of limbless tail- 

 segments within one and the same species. In A. cancriformis the number 

 varies from 5 to 8; in A. numidicus from 10 to 14 ; in A.productus from 4 to 6 ; 

 in A. externum from 5 to 6. This fact is quite in keeping with the undifferen- 

 tiated (that is, embryonic) condition of the posterior region of the body. 



