A DEFINITION OF EVOLUTION 



blood from the common salamander Necturus. If a sample of serum from 

 the immunized guinea pig is then divided among four test tubes, a few 

 drops of serum may be added from Necturus, Amphiuma, Siren, and 

 Cryptobranchus. The greatest precipitate occurs in the tube to which the 

 Necturus serum is added. Amphiuma and Siren sera give about equal pre- 

 cipitates, but rather less than the Necturus serum, while the Cryptobran- 

 chus serum gives only a slight precipitate (Figure 23). On the basis of 

 other criteria (especially comparative anatomy), it is generallv agreed 

 that the first three genera are fairly closely related, while Crijptohranchus 

 is a much more primitive salamander. 



Similarly, if serum from an animal immunized against human blood 

 were divided among five tubes, and serum added from man, an anthropoid 

 ape, an old-world monkey, a new-world monkey, and a lemur, the amount 

 of precipitate formed would also decrease in that order. Thus, the results 

 of serological tests support the theories of relationship which were origi- 

 nally based upon comparative morphology. They indicate that the serum 

 proteins show varying degrees of homology, just as do gross structures. 

 That this should be just a coincidence is not imaginable, yet it is exactly 

 what would be expected on the basis of Darwin's theory that similar 

 species have been formed by descent with modification from a common 

 ancestor. 



The above examples are both taken from among the vertebrates, and 

 indeed this is the group which has been most thoroughly investigated. But 

 extensive studies have also been made on the serological systematics of 

 the Crustacea, Insecta, and Mollusca. Everywhere, the same fundamental 

 result is obtained: animals which had been regarded as closely related 

 on morphological grounds also show close serological affinity. In general, 

 species of a single genus show very close serological similarity; genera of 

 the same family show moderate serological similarity; and families of the 

 same order show slight but detectable similarity. Frequently, different 

 orders of the same class show too little similarity to permit useful com- 

 parisons, but there are exceptions, especially among the birds, wliicli have 

 diverged less serologically than have other classes of vertebrates. 



In general, then, serology has verified the taxonomy which was worked 

 out on other bases, but, in some cases, serological data have been decisive 

 in resolving difficult problems. Tlius the king crab, Xiphosura, was long 

 regarded as a crustacean. Serological tests, liowever, sho^^'ed that it has 

 httle affinity witli typical crustaceans l)ut a strong affinity with spiders. 

 Subsequently, it was demonstrated that these aberrant animals share with 

 spiders the fimdamental arachnidan morphology, while their resemblance 

 to crustaceans is a result of superficial convergence. 



It might reasonably be supposed that only those groups characterized 

 by a blood circulatory system could be investigated by the serological 

 method, liut Mez of Konigsljcrg has successfully applied the method to 

 problems in plant taxonomy. The method is to inject proteins from a plant 

 into a rabbit, thus immunizing the rabbit against the proteins of the plant 

 species used. Antibody-containing serum is then prepared and divided 

 among a series of test tubes. To each is added a few drops of a solution 



58 



