SEPTEMBER 22, 1899.] 
short, that it is one of the rare examples of 
aterm preoccupied as a family name but 
based on a genus which is perfectly valid. 
Nearly eighty years ago Wiegmann * 
proposed the family Glirina for the wom- 
bats (Phascolomys) of Australia and Tas- 
mania, which are now referred to the Phas- 
colomyide. This course was also followed 
by Wagner in 1855, in the Supplement to 
Schreber’s Saugthiere (Vol. V., pp. Xv, 
332). 
In 1887 Ogilby, + in discussing the rela- 
tionships of the peculiar aye-aye( Cheiromys) 
found in Madagascar, remarked: “It is 
only, indeed, the absence of the marsupial 
character which would make us hesitate to 
unite the Cheiromys with the Didelphide ; 
but this cireumstance is so material as to 
require that it should be placed in a differ- 
ent subfamily. At the same time, its 
analogy to the Rodentia ought not to be 
overlooked; and it is for the purpose of ex- 
pressing this relation that I propose to de- 
nominate the small group which I am 
obliged to form for this animal, Gliride. 
I suspect, indeed, that the Cheiromys bears 
a@ more intimate relation to the real dormice 
(Glis) than we are yet aware of.”’ 
Thus Gliridz has been used for three 
different groups of mammals belonging to 
as many different orders, Marsupialia, 
Primates and Rodentia. But since it is 
generally recognized that family names 
must be based on one of the included gen- 
era, this name is not available either for the 
aye-aye or the wombat, while its prior ap- 
plication to these animals invalidates its 
later use for the dormice, the only group 
which contains a genus Gilis. 
It remains to be determined what family 
designation should be applied to the dor- 
mice. Besides Gilisand its synonym Myozus, 
* Wiegmann & Ruthe’s Handbuch d. Zool., p. 52, 
1832. 
Tt Charlesworth’s Mag. Nat. Hist., I., p. 523, Oct., 
1837. 
SCIENCE. 
413 
two other genera have been selected as 
types of higher groups: Platacanthomys, 
made the type of the subfamily Plata- 
canthomyine by Blyth in 1876, and 
Graphiurus, the type of the subfamily 
Graphiurini by Winge in 1887. Either 
of these names might be used for the 
family. Platacanthomyine has the ad- 
vantage of priority, but is open to the ob- 
jection that it represents an aberrant sec- 
tion, so different, in fact, that some authors 
have not associated it with the dormice at 
all. Graphiurus is also aberrant, and ac- 
cording to Winge should be separated from 
all the other genera. This view Thomas 
does not accept, holding that “it might be 
quite as correct to separate Gillis and Mus- 
cardinus on the one side from Eliomys and 
Graphiurus on the other by the pattern of 
the teeth, as to separate the last-named 
from the rest by the structure of the ante- 
orbital region.’”’ Evidently a family based 
on the Indian Platacanthomys or the South 
African Graphiurus would not represent ex- 
actly the same group as that formerly 
known as Myoxidee. ‘ 
Under these circumstances it seems de- 
sirable to adopt a new family name, Mus- 
cardinide, based on Muscardinus, a genus 
which is closely related to Glis. The family 
of dormice may then be subdivided into 
the Muscardinine for Muscardinus, Gli, 
Graphiurus, Eliomys ; and Platacanthomyinee 
for Platacanthomys and Typhlomys, reserving 
Winge’s Graphiurine for Graphirus and 
Eliomys, in case it should be desirable to 
make a third subfamily for these genera, as 
suggested by Thomas. 
T. S. Parmer. 
WASHINGTON, D. C. 
SCIENTIFIC BOOKS. 
Differential and Integral Calculus for Technical 
Schools and Colleges. By P. A. LAMBERT, 
M.A., Assistant Professor of Mathematics, 
Lehigh University. New York, The Macmil- 
lan Company. 1898. Pp. x+245. Price $1.50. 
