SYSTEMATIC PART. 167 



the same reason we cannot point out the exact consanguinity of groups 

 which are apparently related by one or more characters. For we find that 

 when a type starts on a career of development on a specialized line and 

 runs its course, other characters for the time being subordinated to it follow 

 in more or less parallel successions. So that it may come to pass that at 

 a certain time we find two types apparently belonging to different lines of 

 development, which have reached a concurrent condition in some other 

 important character, and we cannot say through which of them the thread 

 of consanguinity has been carried. For instance, in the non -typical Camera ta 

 we have the Platycrinidae without an anal plate accompanied and succeeded 

 by the Hexacrinidae, in which that plate is present. In the typical section 

 the Melocrinidse are followed by the Batocrinidse and Actinocrinidge in 

 similar succession. But (according to our palseontological record) the sym- 

 metric Platycrinidae appear at a somewhat later period than the symmetric 

 Melocrinid^ ; — and as we do not find in the non-typical section any sym- 

 metric predecessors of the Platycrinidge, we would be inclined on this 

 ground to infer that they were derived, from the Melocrinidae. But here 

 we are met by a greater difficulty, for this involves the illogical supposition 

 that the Platycrinidge — a family of the least development of the Camerate 

 type — are derived from one much more highly organized in that particular 

 line. 



The trouble is that all our generalizations are necessarily based upon the 

 Crinoids as they are represented in on?' museums, and not upon the Crinoids 

 as they actually existed in geological time, which is a very different thing. 

 It is like trying to reconstruct a book from detached fragments of the 

 chapters, some of them written in hieroglyphics for whose decipherment the 

 key has not yet been found. We are accustomed to speak of the imper- 

 fection of the geological record, but it is doubtful if in our practical studies 

 we always bear in mind what this really means. To say nothing of the 

 periods antedating the Silurian, in which substantially all vestiges of life are 

 obliterated by metamorphism ; of the accumulations of fossiliferous strata 

 which have been destroyed by erosion during periods of elevation of the sea 

 bottom ; of the strata which over three fifths of the earth are submerged 

 beneath the ocean ; of the great regions unexplored, or covered with ice, 

 snow, or sand ; of the equally extensive areas in which the fossiliferous 

 rocks of one formation are buried under those of succeedino^ ones ; — lea vino; 

 all this out of consideration, how much do we actually know of the life 



