Maech 5, 1915] 



SCIENCE 



367 



a pigeon and more of a water bird than com- 

 monly held? It does indeed, as observers have 

 repeatedly pointed out, have a habitat and 

 habits not unlike those of such water-liking 

 birds as ibises and rails. It inhabits trees 

 and undergrowth along rivers and in marshy 

 regions. It makes nests usually in trees over 

 water. The nests are also, says Beebe, the 

 most recent and most careful observer of the 

 habits of the strange birds, hardly distinguish- 

 able from those of the guinea herons, and 

 built in the same situations. But all this may, 

 of course, mean nothing as to the bird's 

 phylogeny. 



The suggestion that may come from some 

 that my specimens of Lcemohothrium from the 

 hoatzin may have come to this host from some 

 Venezuelan ibis or heron host by natural 

 straggling is extremely unlikely for Mallo- 

 phagan individuals of different bird species. 

 This is only recorded, and practically only 

 possible, among individtials infesting two 

 bird sorts that consort gregariously in con- 

 siderable numbers and closely. This is not 

 true of the hoatzin, as Beebe's observations 

 make clearly evident. Mallophaga are in only 

 rare instances, outside perhaps of crowded 

 hen-houses and chicken yards, colonies of 

 chimney swifts or swallows, and places of 

 common roosting or other foregathering of 

 many bird individuals of a kind, found alive 

 (or even dead) ofE the body of a bird. They 

 make their migration from host individual to 

 individual on occasions of actual bodily con- 

 tact of these hosts, as at mating, and in the 

 nest. 



So it is practically certain that the hoatzin 

 is host to a Mallophagan kind, which is most 

 nearly related to a species, or, perhaps indeed, 

 is but a variety of the very species, found 

 heretofore only on Old and New World ibises 

 and eourlans. 



Vernon L. Kellogg 



Stanford University 



the toxicity of insecticides 

 Certain facts which may be of general im- 

 portance in physiological investigations were 

 brought to light in a study of the toxicity of 



insecticides now under way at the California 

 Agricultural Experiment Station. 



A very elaborate series of determinations 

 were made on the eilect of hydrocyanic-acid 

 gas on scale insect eggs. The plan of the ex- 

 periment was to separate the eggs found be- 

 neath a scale insect into two lots of about 

 equal size, placing them in gelatin capsules, 

 one lot being allowed to hatch without treat- 

 ment, and the other after being exposed to 

 the gas for a definite time. The spiecies stud- 

 ied lay on the average rather more than a 

 thousand eggs, and each series of experiments 

 included the eggs from a hundred insects. 

 Nearly three hundred series were thus studied, 

 including five different species of scale in- 

 sects from eleven different localities in Cali- 

 fornia. 



Solutions of hydrocyanic acid of varying 

 concentration were placed in closed glass 

 containers and the open capsules containing 

 eggs to be treated were suspended above these 

 solutions. The density of the gas above these 

 solutions is dependent on the concentration 

 and temperature. 



After hatching, the capsule was placed 

 under a microscope and an estimate was made 

 of the hatch in each lot, using only the num- 

 bers 05, 10, 20, 30, 40, 50, 60, 70, 80, 90, 95, 

 100 per cent. The following table will show 

 the results of one series. 



The upper right-hand corner gives the re- 

 sults with the weakest dose 'and shortest time. 

 As would be expected, in the opposite corner, 

 there is no hatch and the mean percentages 

 given below show the effects of the different 

 concentrations, the last two or three of which 

 are completely ineffectual since the hatch is 

 the same as the imtreated check lots. 



The series of means given at the right 

 bring out an entirely unexpected result, appar- 

 ently showing that the length of time the ^gs 

 remained exposed to the gas has very little 

 effect. This is, however, not at aU the fact 

 as shown by the curves on the left side of the 

 table. 



The average means of 72 series of experi- 

 ments with the same insect from the same 

 food plant and locality are 58.31, 59.20, 56.10, 



