Ffof. 0. C. Marsh— The Value of Type- Specimens. 549 



The value of a type-specimen, again, may depend largely upon 

 its completeness. Among the invertebrates, especially those now 

 living, types are usually complete enough to show the more 

 important features. This, however, is far from being the case 

 among extinct forms, particularly from the older formations, and 

 the records of Paleontology are burdened with the names of many 

 fragmentary fossils, types of species practically unknown. 



Among the vertebrates of the past, the case is much more serious, 

 and here especially reform in methods is a pressing necessity. 

 From the nature of the case, the older extinct forms are usually 

 represented by fragmentary remains, tlje investigation of which is 

 one of the most difficult problems offered to natural science. A 

 single tooth or a vertebra may be the first specimen brought to 

 light in a new region, and thus become the sole representative 

 of a supposed new form. The next explorer may find more perfect 

 fragments of the same or similar forms, and add new names to the 

 category. A third investigator, with better opportunities and more 

 knowledge, may perhaps secure entire skulls or even skeletons 

 from the same horizon, and thus lay a sure foundation for a 

 knowledge of the fauna. 



As the number of described forms increases, the necessity of 

 a direct comparison of types becomes imperative, and the com- 

 parative value of each type-specimen is thus brought into notice. 

 It will then frequently be found that not a few are uncharacteristic, 

 while others are too incomplete to disclose their own essential 

 features, and hence of little aid in indicating the affinities of forms 

 found with them. 



Type-specimens that do not show characteristic features are, of 

 course, of little value to science, and many such prove a delusion 

 and a snare to the investigator, however faithfully he may endeavour 

 to study them. The imperfect types require still more labour to 

 decipher them. Not a few specimens to-day are types, for the 

 simple reason that they are imperfect. If they had been entire 

 when described, their true nature would have been recognized, and 

 much confusion in nomenclature have been avoided. The chance 

 preservation of some marked features may, indeed, give a hint as to 

 what the whole specimen once was, but too often a suggestion only 

 is thus offered, while the real nature of such types must always 

 remain in doubt. 



A type in Palseontology should consist of the remains of a single 

 individual, and this should stand as the original representative of 

 the name given. A second specimen, or even more, may be used 

 later to supplement the first, but not to supplant it. This, however, 

 has been done by some authors, with the natural result of causing 

 endless confusion in the nomenclature. 



2. The Selection of Type- Specimens. 



The descriptions in Palaeontology are too often descriptive only, 

 and not comparative. This, if well done, is preferable to long 

 academic discussions in regard to the affinities of a specimen of 



