Experiments in Feeding Hummingbirds - 161 



for a very brief season, a large black ant soon taking pos- 

 session and holding the spoils for the rest of the summer. 



The bottles, having been removed from the encroachments 

 of the ants, were ready for the first test. One bird being 

 the sole boarder at that time a level tea-spoonful of sugar 

 dissolved in water was consumed by her daily. In time two, 

 three, four and five Hummingbirds having joined her the 

 quantity of sugar was increased accordingly, a spoonful or 

 two being added to offset any possible waste. In this way 

 more than a pound of sugar was eaten in twenty days, or to 

 be more exact three cupfuls, weighing 9252 grains ; which 

 made an average of 462 grains per day. This for the six 

 birds frequently counted as present confirmed the first rough 

 estimate of a -tea-spoonful of sugar daily for each bird. 



Another method of estimating the amount eaten was de- 

 vised. On several days the sugar and the water were care- 

 fully measured and weighed, then weighed and measured 

 again, after which the syrup resulting from their combina- 

 tion was also measured and weighed, until I felt confident 

 that in a dram of the thinnest syrup served there were forty 

 grains of sugar, or two-thirds of a grarn to every drop. But 

 the syrup usually used was considerably richer than this, 

 easily containing a grain of sugar in every drop ; but it seems 

 best in giving the estimates to keep them to the weakest 

 grade of syrup ever served. 



In making the test a dram of syrup was measured in a 

 glass graduate, and bottle No. 4 was filled. This was always 

 done in the morning when the bottle had been emptied by 

 ants. A waiting Hummingbird came and took her breakfast 

 after which the residue of syrup was poured back intO' the 

 graduate, the bottle being thoroughly drained. Possibly a 

 drop still adhered to the bottle, but the number of minims 

 now in the graduate subtracted from sixty must have given 

 very nearly the amount drank by the Hummingbird. In two 

 summers a number of these tests were made. A bird took 

 for her breakfast from eight to twenty minims, the average 

 being fifteen. Using the low estimate of two-thirds of a 



