194 INDEX GENERUM MAMMALIU M. 
Lagostomidx 1838. Moreover, Chinchillidie, Eriomyide, and Lago- 
stomide are in more or less common use and all apply to the same 
group. It is now known that V7scaccza antedates Lagostomus by 
twelve years; and it can be shown that Zrzomys and Chinchilla were 
published in the same year, but that the latter was more fully described 
and hence is better entitled to recognition; Lagostomus and Eriomys 
being thus reduced to synonymy, Lagostomid: and Eriomyid: need 
not be considered. Of the other two, Chinchillide 1833 was actually 
the earliest family name, whereas Viscachidez 1842 was based on the 
earliest genus. The A. O. U. Code furnishes little help in the solution 
of this question, since it is difficult to say whether Chinchilla or Vis- 
caccia isthe leading genus. Under the Stricklandian Code it is equally 
difficult to determine which is the most typically characterized genus, 
but there is no doubt that V¢seaecia was the earliest known; hence, 
under the second requirement of that code, the family name would 
become Viscacciidee—a term scarcely ever used. 
The American kangaroo rats and pocket mice, comprising the five 
genera Dipodomys, Perodipus, Microdipodops, Perognathus, and 
Heteromys, have long been known under the family name of Sacco- 
myid:, but Saccomys has been dropped as unidentifiable and recently 
Heteromyide has been adopted for the family. At least three of the 
genera, including //eteromys, had been previously selected as types of 
higher groups: In 1853 Gervais named the family Dipodomyna; in 
1868 Gray recognized two tribes, Dipodomyina and Heteromyina, 
and in 1875 Coues separated the subfamily, Perognathidine. Under 
the A. O. U. Code, any one of the three genera //eteromys, Dipod- 
omys, and Perognathus would have claims to recognition as the leading 
genus, though DZpodomys is usually considered the most typical; and 
by selecting the family names according to priority of publication 
and without reference to the date of the genus, Dipodomyide would 
be the proper name. But Allen, in adopting Heteromyide, followed 
the Stricklandian Code literally, basing the family name on the first- 
described genus, although this action transferred the type of the family 
to one of the most aberrant members of the group. 
These examples illustrate the two main difficulties in existing 
rules for family names: (1) Confusion due to the use of several 
names all of which may be applicable to the same group; (2) change 
in type, which sometimes occurs when the family name is uniformly 
based on the first-described genus. Were mammalogists to adopt 
the earliest genus as a basis for the family name such a rule would 
necessitate a number of changes; thus, in the Glires, Caviidee (Cavia, 
1766) would become Hydrocheeridee (//ydrochewrus, 1762); Dasyproctide 
(Dasyprocta, 1811) would become Agoutide (Agoutz, 1799); Erethizon- 
tides (/rethizon, 1822) would become Coendidee (Coendou, 1799), and 
Octodontidee (Octodon, 1832), would become Myocastoridee (J/yocastor, 
1792). 

