52 NORTH AMERICAN FAUNA. 



Quiscalus macrourus Swainson. (ire:it-tailed Grackle. 



Two of these graclcles were shot the latter part of May on a level 

 bit of ground bordering the shore in front of the settlement on Maria 

 Madre. They were the only ones seeu and were undoubtedly strag- 

 glers from the mainland where tliey are abundant and resident near 

 San Bias. 



Astragalinus psaltria mexicanus (Swainson). Mexican Goldfinch. 



Rather common and apparently resident, but nothing distinctive was 

 observed in their habits. On Maria Madre they were usually found on 

 the lower slopes and were most numerous about the settlement. Ten 

 sj)ecimens fail to show any characters distinguishing the island birds 

 from those of the mainland. 



Cardiualis cardiiialis niarise Nelson. Tres Marias Cardinal. 



Cardinalis riryinianus Lawr., Proc. Boston Soc. Nat. Hist., XIV, p. 281, 1871. 

 Ciirdinalis virijinianiis var. igiieiis Lawr., Mem. Boston Soc. Nat. Hist., II, p. 275, 



1874. 

 Cardinalis cardinalis niariw Nelson, Proc. Biol. Soc. Washington, XII, p. 10, 1898. 



Cardinals were very common on Maria Madre and not uncommon 

 on the rest of the group. No one ever molests them, and they were 

 especially abundant about tlie settlement, where they came into the 

 yards and around the houses in the most familiar way. Several pairs 

 could be found at any time during a short walk in the scrubby thickets 

 along the lower slopes of the island. While we were hunting in the 

 low woods it was a common occurrence for them to come very near, and 

 after looking at the intruders with mild curiosity for a short time, to 

 move off through the bushes in quiet j^ursuit of their usual occupations. 

 At other times, while engaged in search of food among the fallen leaves 

 they would scarcely notice one as he walked slowly by within three or 

 four paces. 



Piranga ludoviciana (Wilson). Louisiana Tanager. 



During the first half of May these tanagers were not uncommon near 

 the settlement on Maria Madre, but were not seen on the other islands. 

 Those shot the first of the month were in fair condition, and, several 

 pairs being seen, it was at first considered a resident species. Later, 

 when others were secured, it was noted that they were more and 

 more emaciated, until those killed about the middle of the month were 

 so excessively thin, it was surprising that they had continued to live. 

 About this time the last ones disappeared, no doubt dying from star- 

 vation. From these observations it appeared that the birds must 

 have strayed to the island during migration, about the last of Ai)ril 

 or first of May, and were unable to find a projier food supply. At the 

 same time they feared to start over the sea for an invisible shore and so 

 perished. Another member of the genus, Piraiu/it huh-ntnta flammea, is 

 resident in large numbers on the islands and found an abundant food 

 supply, as was shown by their being among the fattest birds collected 

 during the time that P. ludoriciana was dying of starvation. 



