﻿30 BULLETIN 1364, U. S. DEPAETMENT OF AGEICULTURE 



these traps were fairl}^ satisfactory, for it is believed that the num- 

 ber of dead bees caught by them, before and after the spraying, 

 represented a true mortality index of the bees tested. Because of the 

 difficulties encountered in this kind of investigation, particularly 

 when bees are poisoned but little or not at all, a more reliable method 

 was introduced, in which all the interfering factors were either con- 

 trolled or eliminated. Briefl}^ stated, this method consisted of the 

 following points: (1) The maturing brood was removed from six 

 strong colonies, and by closely following a special procedure while 

 weighing the hives with their contents on the platform scales the 

 daily loss or gain in ounces of the adult bees were obtained inde- 

 pendently of the Aveight of the stores, which were an interfering and 

 variable factor; (2) an undisturbed colony was kept on these scales 

 and was weighed hourly each day; (3) all the adult bees and the 

 maturing brood in two other colonies were weighed three times on 

 these scales, the weight of the stores being eliminated by foUoAving 

 the special procedure just mentioned; (4) 10 outgoing and 10 return- 

 ing field bees caught at the hive entrances were weighed on the 

 chemical balances daily, every hour the bees were working; and (5) 

 daily weather observations were made and recorded. In 1915 the 

 writers worked separately, but thereafter together, using both 

 methods. 



Because of unfavorable weather no definite and conclusive results 

 were obtained at Koswell, IST. Mex., Benton Harbor, Mich., and Win- 

 chester, Va., in 1915. Most of the samples of dead bees collected 

 at Roswell contained arsenic in very small quantities, but no symp- 

 toms of arsenic poisoning were noticed, except in two hives, which 

 contained poisoned pollen collected from cover crops the preceding 

 year. The trees at Benton Harbor bloomed irregularly, so that their 

 blossoms were sprayed in all stages, thus causing a light mortality 

 among the bees, and most of the samples of dead bees contained small 

 quantities of arsenic. There did not seem to be any unusual mor- 

 tality among the bees at Winchester, and not one of the 21 samples 

 collected contained arsenic. (See Tables 2 and 6 for weight records 

 at Winchester.) 



The experiments at Winchester, Va., in 1916, were conducted in 

 three apiaries, surrounded by large commercial orchards. The first 

 apiary consisted of TO colonies, the second of 40, and the third of 

 6. For the weighing experiments the first and third apiaries were 

 used. From the time the trees were almost in full bloom until 

 all the petals had dropped, the weather was almost ideal, and the 

 bees worked normally. When the trees were sprayed nearly all 

 the petals had fallen, and by this time the bees had almost ceased 

 visiting the trees. The results from the various sources did not 

 indicate that the bees tested had been poisoned, although 17 of the 

 28 samples collected contained very small quantities of arsenic, too 

 small to have been fatal to bees. (See Tables 3 and 6 for weight 

 records.) 



The experiments at Fennville, Mich., in 1916, were also conducted 

 in three apiaries, surrounded by commercial orchards. The first 

 apiary consisted of 12 colonies, the second of 14, and the third of 

 30. For the weighing experiments 6 colonies in the first apiary 

 were used. During the entire period of the experiments the weather 



