1S94.] DIDELPni'ID.E OF S.E. BRAZIL. 463 



certainty about the feeding in freedom, the most natural way was 

 to examine the excrements of fresh captured specimens. These I 

 always found to be composed principally of hard remnants of insects 

 and small Arthropoda — elytra of beetles, legs and scales of butter- 

 flies, and wings of flies. Having at my disposal a flourishing brood 

 of meal-worms ( Tenehrio molitor), which I had obtained from Europe 

 as a convenient food for the numerous birds, reptiles, and batra- 

 chians which I have always around me for daily observations of 

 their habits, generally most insufficiently known, it was not very 

 difficult to accustom these marsupials to take the worms. Soon 

 they became very fond of meal-worms and ran to meet the 

 offering hand or pincers. The prey seized, they sit up like a 

 squirrel and so many rodents, and, holding the insect with the 

 hands (opposing sometimes only the first finger, sometimes the first 

 two to the remainder), they crush it rapidly with visible eagerness 

 and audible smacking. This aspect of the graceful animal always 

 reminds me of the European Dormouse. The eyes, like black 

 resplendent pearls, give to the physiognomy of the face an 

 expression particularly confident. All the movements are sudden, 

 rapid, and executed with elegance. The animal is fond of water 

 and milk, and will not delay long when these liquids are offered in 

 a spoon. It drinks often and continuously, lapping like a dog or 

 a cat, and M-ater seems to be a most important article with it. 

 During the day it hkes to sleep in some hiding-place, formed by 

 leaves, cotton, or tow; but the sleep is not very deep, and short 

 diurnal excursions in its cage are frequently observed. It seems 

 to be most susceptible to cold and moisture. Towards the evening 

 the little marsupial becomes more and more lively and agile, and 

 during the night it is more or less in constant movement. There 

 is thus no doubt that its habits are by preference nocturnal, and 

 it is easily comprehensible why these animals are comparatively 

 seldom met with during the day, except by the accidents above 

 mentioned. Nearly all of my prisoners of Micourens pusilhis 

 succeeded in finally escaping during the night ; one was observed 

 for nearly a fortnight after his escape in my study, without any 

 possibility of discovering his hiding-place during the day. He 

 plundered my caterpillars and chrysalids on his nocturnal depre- 

 dations. 



The gait of Micoureus pusillus is somewhat different from that 

 of a rodent of equal size. It is a trot, generally not so rapid as 

 that of a house-mouse. When sleeping the tail is rolled up ; 

 in movement it is extended in a straight line. I have sufficient 

 proofs that Micoureus pusillus is not entirely unable to climb, but 

 I am sure that in general it lives principally on the ground and 

 that it has to be considered as very little arboreal. 



5. Pebamts teistkiatus. 



With certainty I can distinguish only one member of the 

 subgenus Peramys among the material of Didelphyidse collected in 

 the Serra dos Orgaos. I identify it with the P. tristriatvs 



