82 
every  tabulated  part  of  the  epidermis,  offering  about  the  same  type 
as  the  tail  shows. 
The  hands  and  feet  (see  plate  V,  figs.  4  and  5)  are  stouter  deve- 
loped than  in  any  other  rat;  compared  with  Uromys  macropus  the 
measurements  are  for  this  animal  35  and  65  Mm.,  for  M.  armandvillei 
47  and  86  Mm.;  the  differences  in  the  foot-pads  will  be  readily  per- 
ceived by  comparing  Peters'  plate  with  our  plate,  figs.  4  and  5. 
Skull  and  dentition.  In  comparing  the  different  parts  of  the  skull 
(plate  Y,  figs.  1,  2  and  3)  with  the  same  bony  parts  figured  by  Peters 
(1.  c.  figs.  6,  7,  8  and  9)  it  will  strike,  that  although  M.  armand- 
villei is  a  larger  specimen,  its  skull  is  shorter  than  that  of  Uromys 
macropus  and  the  distance  between  incisors  and  first  molar  is  much 
shorter;  the  molar  series,  the  anterior  palate  foramina  and  the  bulla 
auditiva  however  are  much  stouter  and  more  developed  than  in  ma- 
cropus. In  the  latter  the  upper  molar-series  runs  much  more  parallel, 
meanwhile  they  are  conspicuously  convergent  in  armandvillei,  so  that 
the  distance  between  the  first  upper  molars  is  relatively  much  nar- 
rower than  in  any  other  large  Rat  that  I  know.  The  absence  of  an 
outwardly  arched  malar  process  distinguishes  the  skull  of  armandvillei 
at  once  from  the  skull  of  macropus  and  at  the  same  time  from  the 
species  of  mice  belonging  to  the  section  Nesokia,  which  have  a  very 
expanded  malar  process.  Tn  this  respect  our  species  agrees  much  more 
with  the  yellow-tailed  Mus  meyeri,  with  which  species  it  also  has  in 
common  the  form  of  the  lower  molars  (Cf.  my  Catalogue  ostéologique , 
1887,  plate  7,  figs.  5  and  8). 
Appendix  on  Mus  setifer  Horsfield. 
Above  I  mentioned  Mus  setifer  Horsfield,  a  species  always  badly 
understood  by  authors  working  on  Indian  Mice,  see  apud  Blanford, 
Oldfield  Thomas  and  Anderson,  and  this  apparently  by  lack  of  material. 
As  Horsfield  very  exactly  described  the  external  characters  as  well, 
as  remarked  the  transversely  grooved  molars ,  I  think  that  to  the  com- 
plete knowledge  of  the  animal  there  is  only  wanting  a  good  drawing 
of  its  skull.  From  a  study  of  plate  V,  figs.  8,  9  and  10  it  will  be 
evident  that  setifer  has  a  very  expanded  malar  process  and  molars 
with  transverse  laminae  and  that  it  belongs  to  the  section  Nesokia, 
