LIXIfEAN SOCIETY OV LOfOOX. 29 



later tlie cluster becomes compacted, so that, the corallites are 

 elongated and in close contact. Next tlie colony is compressed, 

 so that they become hexagonal in section. Einally, the walls 

 between the individual corallites disappear, and the septa of one 

 are continuous with some of those of its six neighbours. When- 

 ever the complete lineage can be traced this is proved to be the 

 inevitable succession of skeletal forms. 



The iirst lineages to be studied in detail were those of the 

 ammonites, described by the late Alpheus Hyatt. He recognised 

 that as each lineage progressed, and as the successive shells 

 acquired new characters, they invariably passed through the stages 

 represented by their predecessors before they reaclRnl this higher 

 plane. The earlier stages might be abbreviated — some might even 

 be omitted — still there was evidence of them, and they were 

 merely slurred over so that the later adult characters might be 

 hurried on. Subsequent studies of lineages of ammonites by 

 other authors have entirely confirmed Hyatt's conclusions, and 

 the same results have been obtained more recently b}- the 

 examination of many other groups of fossils. 



The tracing of lineages among vertebrates, especially mannnals, 

 lias afforded repeated examples of the recapitulation of ancestral 

 characters in the immature stages of organisms. Sometimes they 

 can only be detected by close observation, but they are neverthe- 

 less evident when poiaited out. Even among tlie scanty fossils 

 representing the lineage of man, there seems tome to be one dis- 

 covery definitely proving a case of recapitulation in the modern, 

 human skeleton, and as this has not hitherto been sufficiently 

 appreciated, I wish to emphasise it by a new diagram (fig. 1). 

 jNIany years ago the late Joseph Leidy pointed out that among the 

 extinct ancestral horses the milk or temporary dentition of a 

 genus always more closely resembled the permanent dentition of 

 its predecessor in the lineage than did the permanent dentition of 

 the same genus. In other words, the temporary dentition of the 

 later and more advanced genus repeated some of the features of 

 the final dentition of the immediately preceding genus. The 

 later type retained the old pattern in its temporary dentition, but 

 progressed forwards in having a more effective grinding dentition 

 for its adult career. If the jjrimitive human skull found at Pilt- 

 down, Sussex, be regarded as belonging to the innnediate 

 predecessor of modern man, exactly the same phenomenon is 

 (ibservable in him. The tem[)orar3" lower canine tooth of modern 

 man very closely resembles in sliape the pern)anent lower canine 

 tooth of the fossil Piltdown man ; and to retain its apex at the 

 level of the crown of the adjacent teeth in the innnature modern 

 jaw it is never completely extruded (fig. 1 a). If tlie base of its 

 crown were raised to the same level as that of the adjacent teeth 

 (fig. 1 b), it would be pronn'nent and interlock with the upper 

 canine to the same degree as in Piltdown man (fig. 1 c) and 

 equally approach the ape-pattern (fig. 1 d). ^Modern man is thus 



