-;iil 
 
 ■■\^ 
 
ALBERT R. MANN 
 LIBRARY ■> 
 
 New York State Colleces 
 
 OF 
 
 Agriculture and Home Economics 
 
 AT 
 
 Cornell University 
 
Cornell University Library 
 QL 767.M11 
 
 The experimental method of testing the e 
 
 3 1924 001 527 294 
 
Cornell University 
 Library 
 
 The original of tiiis book is in 
 tine Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/cletails/cu31924001527294 
 
The Experimental Method of Testing the Effi- 
 ciency of Warning and Cryptic Color- 
 ation in Protecting Animals 
 from their Enemies. 
 
 BY 
 
 W. L. McAtee. 
 
 From the Proceedings of The Academy of Natural Sciences 
 o/ Philadelphia, June, 1912. 
 
 luued September 8, 191B. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 281 
 
 the expeeimental method ot testing the efficiency of wabning 
 
 and cryptic coloration in protecting animals 
 
 from their enemies. 
 
 by w. l. mcatee. 
 Contents. 
 
 Introduction 281 
 
 Experiments with Invertebrates (chiefly Insects) 287 
 
 Experiments with Vertebrates 289 
 
 Fishes 289 
 
 Amphibia 290 
 
 In Countries other than the United States 290 
 
 In the United States 291 
 
 Toads 291 
 
 Frogs 295 
 
 Salamanders 296 
 
 Reptiles 296 
 
 Experiments in Asia 296 
 
 Experiments in Europe 298 
 
 Experiments in America 298 
 
 Mammals 300 
 
 Mixed Groups of Animals 301 
 
 Birds 318 
 
 Experiments in Europe 318 
 
 Experiments in Africa 325 
 
 Experiments in Asia 326 
 
 Experiments in America 331 
 
 Experiments by Judd and Beal 332 
 
 Summary 356 
 
 Introduction. 
 
 The selectionist theories regarding the significance and the causes 
 of production of the so-called warning, mimicking, and cryptic 
 coloration long preceded a knowledge of the food preferences of 
 insectivorous animals sufficient to warrant such speculations. In 
 fact, this knowledge is still almost entirely lacking for many parts of 
 the world — ^including the Amazon valley, which is the home of the 
 brightly colored Heliconiid butterflies and their mimics that sug- 
 gested the mimicry theory to H. W. Bates. In 1861, Bates explained 
 the numerous cases of mimicry of the butterflies of this family by 
 butterflies and moths of at least five other families, by stating that 
 the Heliconiidae probably are unpalatable to insect enemies, and 
 that the others disguised in their dress share this immunity.' He 
 
 1 Trans. Linn. Soc. Lond., XXIII, 1862, p. 510. 
 
282 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 also proposed to explain cryptic resemblances, or the mimicry of 
 inanimate objects by natural selection. Four years later, A. R. 
 Wallace stated^ that he agreed entirely with Bates as to the cause of 
 mimicry, and cited many examples of the phenomenon from the 
 Malayan region. In 1867 the same author first gave a definite 
 theory' concerning the significance of the brilliant colors which are 
 now commonly referred to as warning colors. In 1870* he extended 
 the application of Bates' theory of cryptic and mimicking colors 
 and advanced the results of experiments in support of the theory of 
 warning colors. 
 
 It is not necessary at this time to trace the later development of 
 the theory by Miiller, Dixey, Poulton, and others. Suffice it to say 
 that the original definite suggestion that conspicuous colors have 
 been developed to advertise disagreeable qualities was the result 
 of Wallace's cudgelling his brain at Darwin's instance, for an explana- 
 tion of the coloration of certain insect larvse, which obviously could 
 not be accounted for by sexual selection. The theory has since been 
 expanded to include conspicuous coloration in all groups of animals. 
 A certain insect smells badly to man; is colored red and black, for 
 example, it is conspicuous and nasty to us, hence it must be to 
 insectivorous animals. Its striking color advertises its nauseous 
 qualities and it is avoided after experience; in other words, is pro- 
 tected. So goes the original theory. Although it has been expanded 
 to include all conspicuous forms, whether or not they are nauseous 
 to man, its supporters seemingly find it impossible entirely to forsake 
 the older anthropomorphic ground. Mimicry theories hold that a 
 palatable form gains protection by resembling one of the conspicuous " 
 but nauseous ones, and that distasteful forms are mutually benefited 
 by resemblance. Each of these theories, it may be repeated, was 
 built up in the absence of evidence that the insects concerned were 
 actually distasteful or palatable as claimed. This was the principal 
 criticism made by the comparatively few who at the time dared 
 question the all-sufficiency of natural selection, and it stands to-day 
 the greatest obstacle to acceptance of the theories. 
 
 This criticism spurred the supporters of the theory to sporadic 
 efforts to produce evidence in favor of their contentions. The 
 favorite method of securing such evidence has been by experimenting 
 with captive animals, and the principal body of alleged proof of the 
 
 ^ Tram. Linn. Soc. Lond., XXV, 1866, pp. 19-22. 
 
 ' Proc. Ent. Soc. Lond., 1867, p. Ixxxi. 
 
 * Contributions to the Theory oj Natural Selection, 1870, Chap. III. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 283 
 
 distastefulness of certain insects and the palatability of others 
 consists of the results of experiments. The cry always is to test 
 theories under experimental conditions, but there are many things 
 which cannot be so tested. The very conditions of captivity and 
 the limited choice of food constitute abnormalities which cannot 
 fail to distort the food relations of most animals, and so modify the 
 results of experiments that they bear little or no analogy to natural 
 ■conditions. 
 
 In experiments play is allowed to the fancy of the experimenter, 
 and the interpretation of facial and other expression of the subject 
 has often been given more weight than the actual result of the 
 experiment, that is, whether the insect was eaten or rejected. R. I. 
 Pocock* very frankly admits this, saying, "It is quite clear that the 
 plain record of an insect being eaten is no proof of its palatability. 
 Better evidence on this head is supplied by the behavior of the bird 
 towards it. After a little experience in the matter, I was able to 
 satisfy myself, at all events, as to the approximate correctness of my 
 interpretation of the bird's actions, and to judge thereby of the 
 comparative palatability of the insects they tasted." This is honest 
 confession at any rate, but the writer must take issue with this 
 author as to the value of interpretation of behavior. "The plain 
 record of an insect being eaten," which he holds up to scorn, may not 
 show palatability, but shows something much more definite, namely, 
 that the insect is acceptable food. Palatability in the sense used 
 by some of the experimenters is entirely a figment of the imagination. 
 This is proved by the many cases of refusal in captivity of insects 
 which are eaten under natural conditions, and by the misinterpreta- 
 tion of the following among other features of the behavior of caged 
 birds and other animals. 
 
 Wiping the bill or mouth: If a bird wipes its bill, or a lizard or 
 frog its mouth, when it is being experimented with, the action is 
 almost always credited as a sign of distaste. Yet nothing is more 
 common than to see wild birds wiping their beaks across branches or 
 other objects. It occurs at all times, apparently is often done in a 
 purely mechanical way, and certainly has no essential coimection 
 with the taking of food or perception of tastes. 
 
 Dropping and picking up or in any way manipulating the prey 
 is another thing usually taken as evidence of unpalatability, but 
 nothing could be more at variance with conclusions drawn from 
 
 5 Proc. Zool. Soc. Lord., 1911, p. 810. 
 
284 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 natural behavior. It is highly characteristic of many birds to 
 thresh their prey about, and this is often continued until the object 
 is broken in many pieces, which are separately swallowed. E. B. 
 Poulton considered (P Z. S., 1887, p. 219) that a pupa would have 
 been swallowed whole if palatable, but as it was crushed and the 
 contents eaten it must have been unpalatable; and Weismann 
 {Studies in the Theory of Descent, 1882, Vol. I, p. 341) thinks it readily 
 conceivable that a certain caterpillar may not be unpalatable to 
 lizards, because they swallow it whole, whilst it is perhaps distasteful 
 to birds, because they must hack and tear it in order to swallow it. 
 As a matter of fact, it is habitual for woodpeckers and jays to peck 
 open pupse and extract the contents, and smaller birds, as the chicka- 
 dees and titmice, not only use this method when attacking pupae, 
 but for large larvae also. Many birds hold the food between their 
 feet and hammer it vigorously before eating, and others accomplish 
 the same end by repeatedly picking it up and throwing it down. 
 At least one experimenter, Jenner Weir, recognized such actions as 
 natural, for he says, "All perfect Lepidoptera apparently require 
 preparation before they are swallowed by birds; they are taken 
 between the mandibles, shaken and bruised for a minute or two, and 
 generally have the wings removed before they are eaten."* In many 
 experiments, however, this same action is reckoned as a sign of 
 disgust, if not as an actual rejection. Lizards habitually chew large 
 prey before swallowing, snakes chew it or crush it by constriction; 
 all of these actions are simply part of the normal preparation of food 
 for deglutition, and in no sense of the word evidences of distaste. 
 
 Hesitancy and caution are usually translated as distaste. Does 
 a bird of prey dislike the mouse it holds by a talon on its perch for 
 hours at a time; does a butcherbird dislike the prey it impales on a 
 thorn or wedges in a crotch for future reference? Does a cat play 
 with a mouse because she hesitates to swallow so distasteful a 
 creature? What animal does not employ dilatory tactics in feeding 
 when it is not uncomfortably hungry? This subject naturally leads 
 up to that of disregard, which may be looked upon as hesitancy 
 prolonged. 
 
 Disregard may arise from many conditions, unrelated to the 
 palatability of the food, such as varying appeal of the food to the 
 captive animal according to its state of activity or health, or degree 
 of hunger; or such as the size of the object offered, presence of the 
 
 ' Trans. Ent. Soc. Lond., 1869, p. 22. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 285 
 
 experimenter, or other disturbing elements. Disregard is a frequent 
 phenomenon in experiments with birds of mixed feeding habits, 
 which are most frequently used because more easily provided for 
 in captivity. These birds have no great natural fondness for insects, 
 and certainly not for adults of the order — Lepidoptera — most 
 frequently used in the experiments. The insignificance of disregard 
 is shown throughout Frank Finn's experiments, in the accounts of 
 which {Journ. Asiatic Soc. Bengal, 1895, 1896, 1897), very common 
 expressions concerning cases in which certain butterflies were not 
 eaten in the presence of the experimenter are: "The butterflies 
 were all gone later on," "Next day all the butterflies were gone," 
 etc. In fact, more than 64 per cent, of the butterflies which were 
 left in the cages were eaten in the absence of the experimenter, and 
 more than 77 per cent, of those eaten belonged to the "nauseous" 
 group. The experiments of S. D. Judd (see pp. 332-352) show that 
 disregard bears no particular relation either to acceptances or to 
 rejections, and certainly none to natural preferences. Thus, of the 
 three categories of insects offered to the birds, namely, "warningly 
 colored" insects, others "specially defended," and "non-protected" 
 forms, practically the same proportions (17 to 19 per cent.) were 
 rejected, while the percentage disregarded varied from 3 to 11. 
 
 In Poulton's tabulations of experiments (P. Z. S., 1887) disregard 
 is usually reckoned practically the same as a refusal. Indeed, the 
 original tendency was to consider that disregard showed much 
 greater distaste than any result following trial, for it is pointed out 
 (pp. 193, 194) that the brilliant colors of caterpillars to be of value 
 must generally prevent even trial, because of the fatal consequences 
 to these larvae of very slight wounds. However, Poulton states that 
 out of thirty-seven cases in the "nauseous" group, fifteen were 
 exposed to hungry animals, other food being withheld, and of the 
 fifteen only three remained untasted. Of these two have been 
 shown to be eaten under certain circumstances (p. 225). Poulton 
 estimates disregard properly in one case, that of Lasiocampa quercus 
 larvse. These were disregarded by birds and lizards, and the com- 
 ment on the'evidence is, "Neutral as far as the adult larva is con- 
 cerned" (p. 209). As a matter of fact, disregard is no more of neutral 
 significance in this case than in the fifteen others in which disregard 
 or disregard plus acceptance is the sole evidence upon which proof 
 of distastefulness is claimed. The fact that nine of the eighteen 
 insects of the "unpalatable" group are known to be eaten by wild 
 
286 PROCEEDINGS OF THE ACADEMY OP [June^ 
 
 British birds further suggests the unreliability of disregard as a 
 guide to natural tastes. 
 
 In consequence of the too great fondness of some experimenters 
 for psychological deductions, the writer, in judging experiments, has 
 separated "disregards" for the reasons above stated, and has taken 
 cognizance only of rejections that resulted from actual trial. But a 
 rejection has been credited for each time the experimenter says the 
 object was refused, even if it was eaten later. This is certainly as 
 great a concession in favor of the experiments as can be expected 
 from one who believes not at all in their utility. Tabulations have 
 been made anew (whenever possible) from the detailed accounts of 
 the experiments. It cannot be hoped that these are free from error, 
 but it may properly be assumed that they are approximately correct. 
 
 The earlier experiments especially are characterized by the average 
 small number of tests of the various insects. For this reason, single 
 or a very few acceptances or refusals have been held to prove the 
 palatability or otherwise of a certain form. Thus in the eyes of 
 those who had faith in experiments, results piled up in a really 
 beautiful way. It is not out of order, however, to point out in 
 advance that there are many inconsistencies between the various 
 sets of experiments and that these show that conclusions based 
 upon a few trials are extremely liable to be overturned. 
 
 An interesting case showing the danger of basing conclusions upon 
 a single rejection is given by Jenner Weir {Trans. Ent. Soc. London, 
 1869, p. 22) : "The imago of Spilosqma menthastri .... was tasted 
 by the reed bunting, but not relished, and soon dropped ; the .... 
 bird, however, attracted by the fluttering insect, returned to it, and 
 ultimately swallowed it." If the insect had been taken out, the 
 single rejection would, of course, have stood as the reed bunting's 
 record. A case illustrating the possibilities of single acceptances or 
 rejections is R. I. Pocock's experiment on "the distastefulness of 
 Anosia plexippus" {Nature, 87, 1911, pp. 484-485). A specimen was 
 offered to eleven species of birds and refused a total of thirteen times 
 by them, before it was offered to a tinamou, which swallowed it at 
 once. Reversing the experiment thus might have given an exactly 
 opposite impression of the palatability of this insect. 
 
 In the following pages the expressions accepted, rejected, and 
 disregarded are often abbreviated to A, R, and D, respectively. In 
 general, the experiments are reviewed in groups determined by the 
 classes of animals tested. 
 
i912.j natural sciences of philadelphia. 287 
 
 Experiments with Invertebrates (Chiefly Insects). 
 
 Since Professor E. B. Poulton,'' the most prominent supporter 
 of the theories of warning and mimicking coloration, has come to 
 the conclusion, from proper data — that is, records of insects col- 
 lected with naturally captured prey — that predaceous insects in 
 general are enemies of the "specially protected" groups, it is not 
 necessary to review experiments relating to the food preferences of 
 insects. It is worthy of note, however, that some of these experi- 
 ments have often been quoted as affording support to the prevailing 
 theories. Poulton's latest conclusion is proof, therefore, that even 
 the ardent believers in the experimental method admit that an 
 "ounce" of proof as to natural behavior is worth a "pound" of 
 experimental evidence. 
 
 An annotated bibliography of experiments upon invertebrates is 
 given for the benefit of those who may desire to consult the original 
 accounts. Those entries marked with an asterisk denote experi- 
 ments which were not undertaken with a view to testing protective 
 adaptations. 
 
 Barlow, Edward. A short note on the Food-insects of the Mantis 
 
 Heirodula bipapilla Serv. Proc. Asiatic Soc. Bengal, December, 
 
 1894 (1895), pp. 138-139. 
 Ate Musca, Lucilia, and its own species, and killed but did not eat 
 
 the Hemiptera, Cyclopetia and Physomerus. 
 Belt, Thomas. The Naturalist in Nicaragua. London, 1888. 
 Experiments with Heliconii. "A large species of spider (Nephila) 
 
 .... used to drop them out of its web when I put them into it. 
 
 Another spider that frequented flowers seemed to be fond of 
 
 them, and I have already mentioned a wasp that caught them to 
 
 store its nest with" (p. 317). 
 Butler, A. G. Remarks upon certain Caterpillars, etc., which are 
 
 unpalatable to their enemies. Trans. Ent. Soc. Loud., 1869, 
 
 pp. 27-29. 
 Spiders (Ereiba diadema and Lycosa) rejected larvte of Abraxas 
 
 grossulariata and Halia wauaria (p. 28) . 
 Embody, G. C. [Food of fresh-water Amphipods in aquaria]. Son- 
 
 derabdruck aus Internal. Rev. d. ges. Hydrobiologie u. Hydro- 
 
 graphie, Biol. SuppL, III, 1911 (1912), pp. 4-6. 
 Freshly killed isopods, snails, earthworms, tadpoles, and bits of 
 
 fish, and beef were eaten, but planaria were refused. 
 [Food of Hydra] p. 31. 
 
 Young Hyalella and Eucrangonyx were eaten. 
 Marshall, G. A. K., and Poulton, E. B. Trans. Ent. Soc. Land., 
 
 1902. 
 
 ' See Trans. Ent. Soc. Lond., 1906, pp. 363, 364, 401, 403, and 408. 
 
288 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 Experiments on Mantidse in Natal and Rhodesia (G. A. K. M.), 
 pp. 297-313.. 
 
 [Sununary of these experiments] (E. B. P.), pp. 313-315. 
 
 Conclusions from experiments on Mantidse (E. B. P.), pp. 315-319. 
 
 "We may safely conclude that outside the Acraeinae, and doubt- 
 fully the Danainse, Mantidse devour butterflies very freely, 
 the species with warning colors as well as the others, and that 
 they are far more indiscriminating than the majority of insect- 
 eaters" (p. 316). "But Mr. ^Marshall's experiments yielded 
 plenty of evidence of the positive refusal and acceptance, as it 
 were, imder protest of Acrseinse, so that there can be no doubt 
 of their distastefulness to this class of enemy" (p. 317). 
 
 Despite this conclusion, the fact remains that these Mantidse ate 
 more Acrseinse than they rejected upon trial, and rejected no 
 Danainse. It should be noted, furthermore, that many of the 
 rejections occurred in the last day or two of the life of the 
 various individual mantids when no food whatever was taken. 
 
 Experiments on spiders in the Karkloof (G. A. K. M.), pp. 319-322. 
 Results of experiments on spiders and the earlier experiments on 
 ^Mantidse: one probable meaning of tenacity of life in distasteful 
 " insects (G. A. K. :M.) pp. 322-325, (E. B. P.) pp. 325-328. 
 
 "Experiments have convinced me that both spiders and mantises 
 have no appreciation of warning colors" (p. 322). "It is quite 
 probable .... that certain species of spiders, together with 
 mantides and other predaceous insects, will be found among the 
 chief, perhaps the chief, non-parasitic enemies of aposematic 
 msects" (p. 327). 
 Marshall, G. F. L., and de Xiceville, L. The Butterflies of India, 
 Burmah and Ceylon, Calcutta, 1882-83. 
 
 "Mr. de Xiceville has experimented with the carnivorous Mantis 
 on many of the butterflies believed to be offensive to birds, and 
 he has found .4 . violce is the only butterfly which all the species 
 of Mantis he has experimented with refuse to eat" (p. 318). 
 This is all there is on the experiments. 
 Meisner, Otto. [Food of an Ant-lion and a Clerid in confinement.] 
 Ent. Bl. Xurnberg, 5, Nr. 9, September 20, 1909, pp. 181 and 182. 
 
 A larva of the ant-lion (Myrmeleon) which had been fed only 
 upon flies and caterpillars for a long time, afterwards rejected 
 ants. A Clerus formicarius ate every Adalia bipunctata which 
 got into its cage. 
 *Patch, E. :M. Predaceous Beetles and hibernating Insects. Bui. 
 148, Maine Agr. Exp. Sta., November, 1907, pp. 273-276. 
 
 Pterostichus lucublandu^ was tested with niunerous specimens of 
 Corimelcena pulicaria, Cosmopepla carnifex, Lygus pratensis, 
 and one Serica vespertina, all of which were eaten. Staphy- 
 linidae were tested with the Corimelcena, Lygus, Cosmopepla and 
 cutworms with same result. 
 Peckham, E. G. [Food of spiders and ants in captivity.] Occas. 
 Papers Nat. Hist. Soc. Wis., I, 1889, pp. 107, 109, 110. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 289 
 
 Attidse devoured flies, gnats, larvae and other spiders, but would 
 not touch ants. Synageles picata and Synemosyna formica are 
 always hungry for gnats, but will not eat ants (p. 107). Ants 
 devoured an ant-like spider, Herpyllus, which was placed in a 
 vial with them (p. 109). 
 *Peckham, G. W. and E. G. The Sense of Sight in Spiders with 
 some observations on the Color Sense. Trans. Wis. Amd. Sd., 
 X, 1894-5, pp. 231-261. 
 
 Remarks on food taken in captivity, but nothing in relation to 
 protective adaptations of prey. 
 Plateau, Felix. Observations et experiences sur les moyens de 
 protection de I'Abraxas grossulariata L. Mem. de la. Soc. Zool. 
 de France, Vlt, 1894, pp. 375-392, 3 figs. 
 
 Experiences avec Araignees, pp. 385-388. 
 
 Spiders, Amaurobius ferox, and Agalena labyrinthica would not 
 touch the larva of Abraxas. Tegenaria domestica attacked the 
 larva, but could not pierce its tough skin. Both Tegenaria and 
 Epeira diadema ate imagos, while Agalena rejected them. 
 
 Experiences avec Carabes et Dytiques, pp. 388-390. 
 
 Carabus auratus, Dytiscus marginatus, and D. dimidiatus freely 
 devoured Abraxas larvae. 
 *PococK, R. I. Notes upon the habits of some living Scorpions. 
 Nature, 48, 1893, pp. 104-107. 
 
 They ate cockroaches, blue-bottle flies, etc., in captivity. 
 PococK, R. I. Further notes and observations on the instincts of 
 some common English Spiders. Nature, 49, 1893, pp. 61-63. 
 
 It was found that the prey of Agalena labyrinthica consists largely 
 of bees. A Bombus put in a web was enshrouded before it was 
 killed; a blue-bottle fly was pounced upon at once, while a 
 drone-fly (Eristalis) was cautiously attacked and killed, but not 
 enshrouded. 
 *PoRTEK, J. B. The habits, instincts, and mental powers of Spiders, 
 genera, Argiope and Epeira. Amer. Jouni. of Psychology, 17, 
 1906, pp. 306-357. 
 
 Experiments with food, but not in relation to protective adapta- 
 tions, are described on pp. 334-338. 
 
 Experiments with Vertebrates. 
 
 fishes. 
 
 In a paper entitled, "An Experimental Field-study of Warning 
 Coloration in Coral-reef Fishes,"* Prof. Jacob Reighard records a 
 variety of experiments to determine the significance of colors and 
 flavors of prey to gray snappers {Lutianus griseus) . The predaceous 
 fishes were free and under normal conditions. The common prey 
 
 ' Papers from the Tortugas Laboratory, Carnegie Institution, Vol. II, 1908, 
 No. 9, pp. 257-325. 
 
290 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 of this species at the time of the tests was found to be the sardine- 
 (Atkerina laticeps), a silvery fish that could readily be stained any 
 coIot. a large number of Atherina stained vermilion, yellow, 
 green, blue, or purple were eaten practically as rapidly as fed. Others 
 dyed and treated with formic acid, formaldehyde, red pepper, quinine, 
 ammonia, or carbon bisulphid were taken with equal readiness. 
 Atherina were made unpalatable, however, by sewing in their mouths 
 bits of the tentacles of medusse, and an association of this unpala- 
 tability with a color (red) was established in the individuals of a 
 colony of 150 snappers. The association was found to persist at 
 least five weeks. 
 
 The brightly colored fishes of the coral reefs were then offered tO' 
 the snappers, and they attempted to capture all offered, and actually 
 did take all but one species, of which the single large specimen 
 offered escaped. The species taken were of a variety of colors, 
 including colors and patterns considered as typically warning. In 
 several species " conspicuousness is combined with unpleasant 
 attributes in the form of defensive spines, the typical warning com- 
 bination, yet these fish were all instantly taken" (p. 303). 
 
 It was further found "that the gray snapper discriminates \\\i)x 
 great rapidity and delicacy between the various possible food ele- 
 ments of its environment, which are not conspicuously different 
 from each other," thus proving that the bright colors of the reef 
 fishes would be unnecessary even were their possessors unpalatable- 
 
 Hence "the conclusion is reached that the conspicuousness of 
 coral-reef fishes, since it is not a secondary sexual character and has 
 no necessary meaning for protection, aggression, or as warning, is 
 without biological significance" (p. 320). 
 
 AMPHIBIA. 
 
 In Countries Other than the United States. 
 
 Experiments dealing chiefly with Amphibia are few. Those of 
 Poulton with Hyla^ are cited in another place. A. G. Butler, 
 Eltringham, Plateau, and Finn also record short experiments with 
 animals of this class. Butler published'" the fact that he had found 
 the larvse of Abraxas grossulariata, Halia wauaria, and a sawfly, all 
 fed upon gooseberry, to be distasteful to frogs (and lizards). He 
 asks: "May it not be possible that the plant transmits some pecu- 
 
 » Proc. Zool. Soc. Land., 1887, pp. 269-274. 
 "£n(. Monthly Mag. 5, 1868, pp. 131-132. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 291 
 
 liar acid to the larvae which feed upon it, such as to cause their 
 rejection as food by small reptiles, etc.?" 
 
 Plateau found that Rana temporaria and Triton pundatus rejected, 
 while Triton alpestris disregarded the larvae of Abraxas.^^ Eltringham 
 cites a very few tests made with a salamander. The animal accepted 
 earthworms and honey bees, and disregarded larvae of Pieris 
 hrassicce}^ Finn found that the Indian bull-frog {Rana tigrinay^ 
 took all butterflies offered to it, except two Danais chrysippus, of 
 which species it ate one. 
 
 In the United States. 
 
 A number of experiments upon Amphibia have been performed in 
 the United States. These are discussed in two groups, those on 
 toads»and those on frogs. 
 
 Toads. — In company with Dr. A. K. Fisher, the writer once 
 attempted to give a toad {Bufo lentiginosus) his fill of hymenopterous 
 food. This occurred on Plummer's Island, Md., about 1905. Honey 
 bees and wasps of the genera Polistes, Sceliphron, and Vespa were 
 captured, their wings were clipped, and they were put down so that 
 they would crawl in front of the toad, which was partially domesti- 
 cated about the cabin and was not much disturbed by the move- 
 ments of humans. The toad took every insect offered, although at 
 times he showed considerable but ludicrous signs of discomfort. Not 
 less than 30, and perhaps as many as 40 Hymenoptera were taken by 
 this animal in about an hour. He finally left the spot, apparently 
 to get away from a locality characterized by such extremely spicy 
 food, which nevertheless he was apparently unable to refuse. 
 
 In another experiment performed by the writer at the same locality, 
 on August 6, 1911, another toad was also fed Polistes, Pelopoeus, and 
 another stinging wasp, none of which was refused. A sphinx moth, 
 a small white moth, several ants and flies also were taken. The 
 toad attempted to eat a katydid {Cyrtophyllus perspicillatus) , but 
 found it too large. A small Heterocampa larva, colored green and 
 red, was eaten, and then a Julus was put before him. As soon as 
 it began to crawl he seized it by one end and, not getting a good hold,, 
 ejected it. It then crawled over his head. This might be construed 
 by some as evidence of dislike, but I think that if the Julus had been 
 fairly seized it would have gone down. At any rate the case well 
 
 " M&m. de la Soc. Zool. de France, VII, 1894, pp. 383, 384. 
 « Trans. Ent. Soc. Land., 1909, pp. 473, 474. 
 ^^Journ. Asiatic Soc. Bengal, 66, 1897 (1898), p. 533. 
 
292 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 illustrates the danger of drawing conclusions from scanty experi- 
 menting, for stomach examinations show Julus to be a favorite food 
 with toads. A. H. Kirkland says in his valuable paper on the 
 economic value of the toad:" "Myriapods form a constant article 
 of diet for the toad. Species of the genus Julus were present in the 
 majority of the stomachs examined, the largest number found in a 
 single stomach being seventy-seven. These creatures form 10 per 
 cent, of the food for the season." 
 
 Mr. Kirkland's examinations of .stomachs serve to check another 
 set of experiments which is claimed to show that squash bugs (Anasa 
 tristis) are seldom eaten by toads. The account of these experiments 
 by C. M. Weed and Albert F. Conradi is as follows : 
 
 "The common toad has been generally considered an enemy of 
 the squash bug, being frequently referred to in this connection in 
 articles concerning the pest. We made a large number of observa- 
 tions on this phase of the subject, the most interesting result being 
 the discovery that the odor given off by the bugs will actually kill 
 toads if confined in a small open vessel, such as a wide-mouthed 
 bottle. Some of these experiments as recorded. in a published letter 
 by Mr. Conradi are as follows:'* When a squash-bug nymph of the 
 fifth stage was suddenly introduced into a half-pint, open, wide- 
 mouthed bottle containing a half-grown, live toad, so that the 
 batrachian would get the full effects of the pungent fumes secreted 
 by the bug, the toad was thrown into a temporary stupor, the effect 
 being similar to that of chloroform. As the number of bugs was 
 increased, the effect on the toad was increased. When as many as 
 seven bugs were introduced, the toad fell into a profound stupor, from 
 the effects of which it died in the course of twenty-four hours. 
 
 "On September 8, an adult toad that had been kept in the labora- 
 tory vivarium with a scant food supply for several days, was placed 
 in a quart jar of the same construction as the one mentioned above, 
 and eight bugs were introduced; these bugs, however, had been so 
 much disturbed previously that the source of the pungent secretion 
 had been temporarily exhausted. The toad hesitatingly devoured 
 three, after which she would remove with her front feet every speci- 
 men that made an attempt to ascend the wall of her enclosure; but 
 these bugs were not eaten. The toad was then transferred to another 
 jar of the same size and construction, and eight bugs were suddenly 
 
 i< Bui. 46, Hatch Exp. Sta., 1897, p. 15. 
 
 ^^ Science, N. S., Vol. XIV, No. 360, November 22, 1901, pp. 816, 817. See 
 also Science, N. S., Vol, XIX, No. 479, March 4, 1904, pp. 393, 394. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 293 
 
 introduced from the squash leaf so that the animal would get the 
 first and fullest effects of the odor; the result was that the toad went 
 through a series of contortions followed by a short period of stupor 
 similar to that mentioned before. Upon recovery the toad was 
 again removed to the vivarium, where it now lives in partial hiber- 
 nation. 
 
 "A young, red-spotted salamander was affected and killed as 
 easily as the half-grown toad, while for the common field frog a 
 greater number of bugs were required to bring about similar effects, 
 the frogs also being killed. Many experiments with snakes were 
 tried, but no ill effects from the secretion of the bugs were apparent. 
 
 "The odor that the bug secretes is contained in a clear, slightly 
 greenish liquid expelled from the extremity of the alimentary canal; 
 when it comes in contact with the air the odor is given off almost 
 instantaneously while the liquid remains to evaporate. 
 
 "Further observations showed that toads in confinement would 
 eat squash bugs when very hungry, but we do not think that toads 
 ordinarily devour many of the pests. "'* 
 
 It is most obvious that the conditions of these experiments are 
 never even faintly simulated under natural conditions. The con- 
 clusions in the last paragraph, being based on the results of the 
 experiments, are therefore unwarranted. Moreover, they do not 
 agree with the statements of other observers relating to the habits 
 of the toad under normal conditions. Kirkland found Anasa trisiis 
 in collected stomachs," as did also Judd and the writer. 
 
 Kirkland briefly records an experiment of his own as follows: 
 
 "The writer once confined for study a large toad in a shaded 
 out-of-door box filled with damp earth. To provide suitable and 
 sufficient food for it was quite a task until an entirely satisfactory 
 expedient suggested itself. A hard bread-crust was soaked in 
 molasses and placed in the cage. Bees, wasps, ants, flies, and beetles 
 came to this bait, and it was most interesting to watch the toad 
 seize the flying insects, often before they had alighted on the bread. 
 Stinging insects, bees, wasps, etc., when swallowed by the toad 
 apparently produced uncomfortable sensations for a short time. 
 Fish-worms when captured by the toad often prove too much to be 
 swallowed at once, and when this is the case the fore limbs are brought 
 
 '' "The Squash Bug," Bull. 89, New Hampshire Agric. Exp. Sta., February, 
 1902, pp. 21-23. 
 
 " Bui. 46, Hatch Exp. Sla., 1897, p. 26. 
 
294 PROCEEDINGS OF THE ACADEMY OF [JunC, 
 
 into use to force the unfortunate worm into the capacious gullet 
 of its captor. "'* 
 
 In this connection we may quote C. F. Hodge," who, in giving 
 instructions as to the use of toads as insect catchers in houses, says : 
 "Sugar solutions should not be used [as bait] on account of the 
 danger to honey bees which a toad will take in great numbers despite 
 their stings." Hodge quotes'" a feeding test by Miss E. M. Foskett, 
 the insect used being Macrodadylus subspinosus. Miss Foskett 
 says: "One day I gathered a quantity of rose bugs in a tin box, 
 sat down in the shade beside my queer pet and began feeding bugs 
 to him. At first I did not count, but finding his appetite so good, 
 I started to count. When I had counted over eighty bugs and the 
 toad showed no signs of wishing to conclude his meal, I picked him 
 
 up Previous to my beginning to count, he had taken 
 
 anywhere from ten to twenty bugs. He was quite a large toad, but 
 the bugs were large, too, and very 'scratchy.'" 
 
 The American Sportsman (Vol. 3, No. 2, October 11, 1873, p. 23) 
 reports a series of experiments with toads by Dr. Thomas Hill. 
 This account does not have a thoroughgoing appearance of verity, 
 but this may be not a reflection on the experiments themselves, but 
 upon the reporting. It is said a toad ate "yellow-striped" locusts, 
 earthworms, and at one meal twenty-three squash bugs and ninety- 
 four larvae of Pygcera menistra [Datana ministra]. 
 
 C. V. Riley briefly states" the results of offering larvse of Anosia 
 plexippus to various animals, as follows: "Prompted by ... . 
 experiments made in England, I was led to make similar ones with 
 our gayly colored Archippus larva, and the result fully accords with 
 that obtained by Mr. Weir, for neither turkeys, chickens, toads, nor 
 snakes would touch it." 
 
 Included with some notes on the Florida chameleon {Anolis 
 principalis), Dr. S. Lockwood records^' an observation upon captive 
 toads. Two of these animals ate, respectively, three and two potato 
 beetles {Leptinotarsa decemlineata) , after which they would take no 
 more. Dr. Lockwood then remarks: "It was specially observable 
 of the one which had swallowed the three spearmen, despite the 
 grotesque gravity of his demeanor, that there was a certain dolorous 
 
 "L.c, p. 11. 
 
 " Nature Study Leaflet, Biol. Ser. I, Worcester, Mass., 1898, p. 11. 
 
 2" L. c, p. 10. 
 
 2' Third Ann. Rep. on the Insects of Missouri, 1871, p. 148. 
 
 « Am. Nat., 10, 1876, p. 8. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 295 
 
 air about him, as of one suffering from an overdose of Doryphora. 
 Though kept some two weeks with no other food, neither Bufo would 
 touch a spearman again." 
 
 No better illustration could be asked of the misleading character 
 of experimental results nor, it may be added, of the highly imagina- 
 tive conclusions drawn therefrom. Notwithstanding the "dolorous 
 air" of these toads by reason of potato-beetle diet, the fact remains 
 that toads habitually feed on potato beetles. Tenney,^' Riley,^'' 
 Kirkland,^^ Garman,^^ and Chittenden" among others record toads 
 as enemies of potato beetles. The writer has found the remains of 
 no fewer than twelve Leptinrtarsa decemlineata in a single casting 
 of a toad. 
 
 Frogs. — In an article entitled, "Habit Formation in Frogs,"-' 
 A. A. Schaeffer says: "Individuals of three different species of 
 frogs, Rana clamata, R. sylvatica, and R. virescens learned to avoid 
 disagreeable objects, such as hairy caterpillars, in from four to seven 
 trials or possibly less. Such habits persisted for at least ten days, 
 but this point was not thoroughly tested. 
 
 A Rana clamata formed a habit of avoiding earthworms treated 
 with chemicals in two trials. This habit persisted perfectly for only 
 a short time, covering five trials in about twenty-two hours. The 
 habit persisted somewhat imperfectly for five days. After an electric 
 stimulus had been applied, earthworms were not eaten for seven days, 
 although mealworms were eaten" (p. 334). 
 
 These observations point to the conclusion that any color may 
 be regarded as warning, provided a sufficiently disagreeable impres- 
 sion becomes associated with it. 
 
 Another experimenter, Charles W. Hargitt, was led to doubt the 
 quick formation of such associations by Hyla, as is shown by his 
 ■comments on the behavior of a tree frog toward Hymenoptera. In 
 food taking, he says: "Hyla behaves quite similarly to others of 
 its kind. It seems not to notice any except moving objects. A 
 spider may remain quietly in a given part of the cage for hours or 
 days undisturbed. If it assume an active attitude it is almost 
 certain to be taken very promptly. It is thus with any prey. The 
 insects most commonly supplied were flies, small beetles, grass- 
 
 » Amer. Nat., 5, 1871, pp. 170, 171. 
 
 " Fourth Mo. Rep., 1872, p. 16, and in many other publications. 
 
 " Bui. 46, Hatch Agr. Exp. Sta., 1897, p. 25. 
 
 " Bui. 91, Ky. Agr. Exp. Sta., 1901, p. 66. 
 
 " Circ. 87, U. S. Bur. Ent., 1907, p. 12. 
 
 "-'Joum. An. Behavior., Vol. 1, No. 5, Sept.-Oct., 1911, pp. 309-335. 
 
296 PROCEEDINGS OF THE ACADEMY OF [Juue, 
 
 hoppers, spiders, etc. On one occasion a small wasp was released 
 in the cage and at once began to buzz about or run up the sides of 
 the cage actively. It was but a few moments ere a specimen leaped 
 eagerly and captured the prey. Then a most interesting performance 
 took place. No sooner was the wasp seized than it was whipped 
 into the mouth, and in turn stung the frog; the frog in turn showed a 
 very lively appreciation of that fact, and made an apparent effort 
 to eject the creature; but the process of ingestion had gone too far 
 and deglutition was completed without further ado, nor did the 
 frog show the least further sign of distress. On another day the 
 operation was repeated and very much after the fashion of the 
 preceding. It may be doubted whether Amphibia show any particu- 
 lar discrimination based on that type of experience."^' 
 
 Salamanders. — Albert M. Reese, in a paper on the "Food and 
 Chemical Reactions of the Spotted Newt, Diemydylus viridescens," 
 says that the animals show no difference in reaction toward bits of 
 raw meat and earthworms nor to the juices from these substances."* 
 
 A specimen of hellbender (Cryptobranchus allegheniensis) which 
 ejected remains of a crawfish soon after capture, refused to eat any 
 of these animals placed in its aquarium later.'* 
 
 Reptiles. 
 Experiments in Asia. 
 
 Among Frank Finn's many experimental contributions to the 
 theory of natural selection are two which deal with the food taken 
 by lizards. The first'^ deals with the Indian lizard {Calotes versi- 
 color), both captive and free individuals of which were offered a 
 variety of adult Lepidoptera. The results of feedings of the free and 
 confined individuals agree very well except in the case of butterflies 
 of the genus Euplcea. The record for these insects with lizards in a 
 cage is A 4 Rl, and with those unconfined, Al R4. Species of 
 Danais, Delias eucharis, and Papilio aristolochice were freely eaten, 
 and Finn concludes: "The behavior of these reptiles certainly does 
 not appear to afford support to the belief that the butterflies, at any 
 rate, usually considered nauseous, are distasteful to them" (p. 48). 
 
 The second series of experiments we refer to are reported in 
 
 2» "Behavior and-Color Changes of Tree Frogs," Journal of Animal Behavior, 
 Vol. 2, No. 1, Jan.-Feb., 1912, pp. 53, 54. 
 
 " Journ. Animal Behavior, Vol. 2, No. 3, May- June, 1912, p. 207. 
 ""Oconomowoc," Forest and Stream, 8, No. 20, June 21, 1877, p. 320. 
 ^ Journ. Asiatic Soc. Bengal, 65, 1896 (1897), pp. 42-48. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 297 
 
 Natural Science, December, 1892.33 They deal chiefly with the East 
 African lizards, Mabuia striata and Hemidactylus mabuia. Several 
 presumably distasteful insects were refused by these species. They 
 refused wood lice also, which were eaten, however, by Gerrhosaurus 
 major. 
 
 Experiments to determine the tastes of Calotes were also per- 
 formed by R. C. Punnett in Ceylon.^^ Punnett concludes that the 
 lizard tested (in confinement) by him showed no discrimination in 
 the choice of various adult Lepidoptera offered. "The presumably 
 distasteful Danais was eaten before the presumably palatable 
 Euschema or Mycalesis, and the so-called distasteful Euploea was 
 taken before the supposedly palatable Junonia iphita of not verj' 
 dissimilar coloration. Nor was any hesitation manifested towards 
 Papilio aristolochioe with its postulated evil taste and marked warning 
 coloration" (p. 13). Punnett also found the larvse of the last-named 
 insect as well as an adult Danais plexippus were eaten by another 
 lizard (Lyriocephalus) . "From such experiments as these one can 
 hardly fail to draw the conclusion that Calotes as well as Lyriocephalus 
 will readily eat anything in the way of butterflies that they come 
 across. Nor is this surprising, in view of the fact that such noxious 
 creatures as the large ant {(Ecophylla smaragdina) and hairy cater- 
 pillars constitute a considerable proportion of the contents of their 
 .stomachs. They certainly do not appear to exercise that nice 
 discrimination with regard to butterflies, which is necessary for the 
 establishment of mimicking forms on the theory of natural selection" 
 
 (p. 13). 
 
 Lieut. -Col. Neville Manders also performed experiments with 
 lizards'" in Ceylon, using the following species: Calotes versicolor, 
 C. ophiomachus, C. nigrilabris, C. zeylanica, and Ceratophora stoddarti. 
 The last two species are smaller than the others and would not try 
 to eat butterflies either when caged or free. The experiments with 
 free specimens of the other three species resulted as follows: Lepi- 
 doptera classed as edible. All; Lepidoptera classed as nauseous, 
 A 17 R 3. Manders says one of the latter was at first refused because 
 of large size, then partly eaten, and the other two were too dry. 
 
 He further states "that so long as the butterflies remained per- 
 fectly still, they were entirely unnoticed by the lizards, though they 
 might be in close proximity to them" (p. 708). This indicates that 
 
 3s I, No. 10, pp. 746, 747. 
 
 3* Spolia zeylanica, VII, Pt. XXV, September, 1910, pp. 12, 13. 
 
 35 Proc. Zool. Soc. Land., 1911, pp. 707-710. 
 
298 PROCEEDINGS OF THE ACADEMY OF [June^ 
 
 color is not the stimulus to capture, but that motion is. Manders 
 concludes: "It would seem that those who assume that reptiles 
 take no part in the production of Batesian or Mtillerian mimicry 
 are correct, though further experiments are required" (p. 710). 
 
 Experiments in Europe. 
 
 A series of experiments by H. Eltringham with Lacerta viridi^^ 
 have as their main point the demonstration that a certain lepidop- 
 terous larva {Boarmia rhomboidaria) , was more distasteful to the 
 captive lizards when it had fed on ivy than when fed on apple. In 
 addition, a number of other insects and other invertebrates were 
 used in the experiments. No general conclusions are given. Possible 
 comparisons with Poulton's tables of experiments (1887) are as 
 follows : 
 
 Poulton's tables, 
 
 1887. Eltringham. Animal tested. 
 
 Pieris brassicoe, larva R. A 4 D. Lacerta viridis. 
 
 Apis mellifera A many. R. " " 
 
 Later, Eltringham says that the caterpillars referred to in his 
 previous experiments are not Boarmia rhomboidaria, but Odontoptera 
 bidentata. The adult moths were eaten by lizards, to which they 
 were fed by Messrs. Eltringham and Pocock. The latter found that 
 they were acceptable also to a bird, Graculifera melanoptera. Eltring- 
 ham concludes that the distastefulness of the larvae was not intrinsic, 
 but due to the character of the food in their digestive tracts.'^ 
 
 Plateau states'* that seven Cistudo europasa ignored the caterpillar 
 of Abraxas grossulariata, while one tried and rejected it. Coluber 
 oesculapii and Lacerta muralis disregarded the larvae and rejected 
 them when placed in their mouths. 
 
 Experiments in America. 
 
 Few experimental tests of the efficiency of the protective adapta- 
 tions of insects have been made in the United States. The most 
 important series hitherto published in full was performed by Annie 
 H. Pritchett and was reported in the Biological Bulletin (Vol. 5, 
 pp. 271-287, 1903). The animals used were Sceloporus floridanus,. 
 Gerrhonotus infernalis, Crotaphytus collaris, Cnemidophorus sexli- 
 
 M Trans. Ent. Soc. Land., 1909, pp. 471-478. 
 
 " "Edibility of Lepidopterous Larvae, " Proc. Ent. Soc. Land., 1910, pp. xxxi, xxxii.. 
 
 » Mem. de la Soc. Zool. de France, VII, 1894, p. 383. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 299 
 
 neatus, Eumeces sp., and Phrynosoma cornutum. A large variety of 
 invertebrates were offered the lizards, with the following principal 
 results: (a) Only one instance of a lizard eating a dead insect; 
 (6) insects that move slowly do not attract the attention of the 
 lizard so much as do the more active forms, hence those that remain 
 quiescent are rarely even attacked; (c) insects below a certain size 
 are apparently not perceived by the large species of lizards; (d) 
 large beetles having hard elytra are seldom eaten; (e) the myriapod 
 /mZms was not eaten by any lizard; (/) although the combinations of 
 black and yellow, black and orange, or black and red are supposed 
 to serve the purpose of warning coloration, all insects possessing 
 these colors were, at one time or another, eaten, with the possible 
 exceptions of Panorpa nuptialis Gerst. and a malodorous Lygaeid bug. 
 Some experiments with Sceloporus undulatus by Dr. S. D. Judd 
 which have never been published may now be put on record for 
 purposes of comparison with the series just noted. The results of 
 these tests follow : 
 
 Orthoptera : 
 
 Blattid^ — Stylopyga orientalis (black) A 2 
 
 Mantid^ — Stagomantis Carolina (dark brown) R 
 
 LocusTiDiE — Microcentrum sp. (green) R 
 
 GryllidjE — Gryllus sp. (dark brown) A 3 
 
 Coleoptera: 
 
 Carabid^ — Carabidw (undet.) A 
 
 Harpalus pennsylvanicus (black) A 
 
 Scarites subterraneus (black) R 
 
 CocciNELLiD^ — Coccinella sp. (warning colors) R 2 
 
 Hippodamia sp. (warning colors) R 
 
 Adalia sp. (warning colors) D 
 
 Epilachna borealis (yellow and black) A 
 
 Dermestid^ — Dermestes sp. (nearly black above, 
 
 white below). A 
 
 LampyridjB — Chauliognathus sp. (warning colors) R 
 
 ScarabjEIDjE — Ligyrus sp. (reddish-brown) R 2 
 
 Lachnosterna sp. (reddish-brown) R 2 
 
 Chrysomelid^ — Leptinotarsa decemlineata (yellow and 
 
 black) R2 
 
 Diabrotica sp A 
 
 Heteroptera (all strong smelling) : 
 
 PENTATOMID.E — Brochymena sp R 
 
 Nezara hilaris (green).., R 
 
 REDUViiD.ffl — Arilus cristatus (red-brown, wings with 
 
 bronzed tips) R 
 
300 proceedings of the academy of [junc, 
 
 Lepidoptera: 
 
 ArctiiDjE — Hyphantria cunea 1. (yellow, brown, and 
 
 black, very hairy) A2. R2 
 
 LiPARiDiE — Orgyia leucostigma 1. (red, black, white, 
 
 and yellow, hairy, tufted) R 2 
 
 Diptera: 
 
 MxJsciDJS — CaUiphora erythrocephala (metallic blue) D 
 
 Sceloporus floridanus and S. undulatus are very close relatives; 
 hence so far as the lizards are concerned, comparsions between the 
 two sets of experiments are not far-fetched. Unfortunately, the 
 insects offered have nothing more than the genus in common, and 
 that in only a few cases. However, the comparisons possible on 
 this basis are given: 
 
 Pritchett. Judd. 
 
 Brochymena D R 
 
 Chavliognathus A 24 R 2 R 
 
 Harpahis A 3 D 4 A 
 
 Gryllus ".. A 5+ A3 
 
 Summing up Judd's experiments, we have the following results: 
 
 A. 
 
 "Protected" group" 4 
 
 "Non-protected" group. 8 
 
 Evidently these Sceloporus were hard to please, accepting barely 
 more than a third of all the insects offered; 66 per cent, of the 
 individuals of "protected" species were refused and 52 per cent, of 
 the "non-protected" group. 
 
 Dr. S. Lockwood briefly records** the food habits of the Florida 
 chameleon {Anolis principalis) in captivity. The lizards ate flies 
 and spiders, but would not take the potato beetle (Leptinotarsa 
 decemlineata), and, in fact, were not fond of beetles at all. Dr. Lock- 
 wood cites an observation by Bell, that a pet Anolis, catching an 
 Epeira diadema by the leg, was bitten by its captive and death soon 
 ensued. 
 
 MAMMALS. 
 
 Firm records" brief experiments with an East African mongoose 
 (Crossarchus fasdatus). This animal refused one specimen of a 
 frog {Xenopus lavis) and ate and vomited another. It ate and 
 
 " That is those "warningly" colored or otherwise "specially defended " 
 
 « Am. Nat., 10, 1876, pp. 7, 8. 
 
 «' Natural Science, I, No. 10, December, 1892, pp. 746-747. 
 
 R. 
 
 D. 
 
 10 
 
 1 
 
 10 
 
 1 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 301 
 
 vomited parts of a lizard {Mabuia striata), but the lizard was never- 
 theless entirely eaten. The mongoose was unwilling to eat birds 
 and refused to attack a conspicuous milliped (Spirobolus) . An 
 obscurely colored milliped also was refused by a lemur {Galago). 
 
 Another of Finn's experiments concerning the tastes of mammals 
 for insects deals with a tree-shrew {Tupaia ferrugiiiea).*^ The 
 conclusion is: "It is obvious that this animal had a very strong 
 objection to the 'protected' Danainae and Papilio aristolochice, as 
 it so constantly refused them" (p. 532). This is a fair summary of 
 the experiment except as it applies to Papilio aristolochice, the record 
 for which was A2 R2. 
 
 Marshall and Poulton have published*^ accounts of experiments 
 with a mongoose {Herpestes galera), baboons, and a inonkey (Cerco- 
 pithecus pygerythrus) . The mongoose tested by Marshall (pp. 376- 
 378) refused but one insect consistently and had only two trials 
 with that. The animal was tested with birds as food and refused 
 five out of ten kinds offered. Two of the five refused have colors 
 of the type called warning and this is peg enough upon which to hang 
 some speculations as to distastefulness. Nothing is said about the 
 equally conspicuous colors of two of the species eaten, viz., Nettopus 
 auritus, blackish-green, white and rufous; and Saxicola pileata, 
 chestnut, black, and white. The results of single trials of several 
 insects with Cercopithecus pygerythrus are recorded on p. 379, and 
 pp. 380-392 are devoted to an account and discussion of more 
 extended experiments with baboons. Poulton tabulates the Cole- 
 optera accepted and rejected by the baboons, and from these tabula- 
 tions it appears that about 75 per cent, of the beetles rejected had 
 warning color patterns, as did about 55 per cent, of those accepted. 
 It is unfortunate that there are no records of the natural food of 
 these African mammals that can be used as a check on the experi- 
 mental results. 
 
 MIXED GROUPS OF ANIMALS. 
 
 Brief notes on experiments with a marmoset and lizards are 
 included in E. B. Poulton's description of the "means of defence 
 adopted by the larva of Stauropus fagi" and the " defensive value of 
 'tussocks' of Orgyia and the associated black intersegmental mark- 
 ings. "« 
 
 « Journ. Asiatic Soc. Bengal, 66, 1897 (1898), pp. 528-532. 
 
 « Trans. Ent. Soc. Lond., 1902, pp. 376-392. 
 
 " Trans. Ent. Soc. Lond., 1888, pp. 581-588 and 589-591. 
 
302 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 We are told that when at rest the larva of Stauropus fagi resembles 
 a withered beech leaf. Next, it is stated that the second and third 
 pairs of thoracic legs are so held that they resemble, " in the most 
 beautiful manner, a bunch of brown scales (the stipules of the foliage 
 leaves) which enclose the buds of the beech, and hang down after 
 the latter are unfolded." As if withered beech leaves, thus adorned, 
 were common during the life of this larva and as if beech were the 
 only food plant of the larvae. As a matter of fact, Stauropus feeds 
 on the foliage of several other woody plants. When aroused the 
 larva is said by Poulton to assume a terrifying attitude, the main 
 suggestion of which is a spider-like creature. 
 
 It is rather amusing to contemplate the variety of resemblances 
 claimed for Stauropus larvae. For instance, Poulton himself, ten 
 years later, insists upon an entirely different resemblance from those 
 above mentioned. He then says : 
 
 "The young larvae of Stauropus fagi have often been described as 
 resembling ants. The likeness has recently been analyzed in much 
 detail by Portschinski {Coloration marguante et Taches ocellees, 
 V, St. Petersburg, 1897, p. 44). This acute observer considers that 
 the head of the larva represents the globular abdomen of the ant, 
 while the head and antennae of the latter are suggested by the larval 
 caudal shield with its two appendages. He believes the disturbed 
 larva represents an ant which has seized and is endeavoring to carry 
 
 off some object on the branch which it is exploring During 
 
 the present summer (of 1898) I have had the opportunity of studying 
 these larvae. The young larvae were thought to be ants by all the 
 friends to whom they were shown. One lady considered that they 
 were 'double ants' — an interpretation evidently due to their dis- 
 proporCionate length and to the head-like appearance of the caudal 
 shield. "« 
 
 Birchall^* states that the young larva closely resembles a twig of 
 beech with unopened buds, and that when feeding its likeness to a 
 great earwig or to a Staphylinus is very striking. He also remarks 
 upon the general suggestion of a crustacean in the larva's aspect, 
 but he further desires "to speak doubtfully of the sharp eyes of a 
 bird or Ichneumon being deceived when engaged in its own special 
 business, by any such colorable imitation" (p. 233). 
 
 Mrs. Bazett" notes the great resemblance that the newly hatched 
 
 « Journ. Linn. Soc. Land., Zoology, Vol. XXVI, 1898, pp. 589, 590. 
 « Enl. Monthly Mag., XIII, 1877, p. 231. 
 " Ent. Rec, II, 1891, p. 210. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 303 
 
 larva bears to an ant, while Kirby^ says it is from the extraordinary 
 appearance of the larva that the species derives its name of lobster 
 moth. 
 
 Thus the larva of Stauropus is supposed to mimic more or less 
 closely, objects in both the vegetable and the animal kingdoms, 
 and within the limits of the latter, representatives of five orders 
 (not to mention ants carrying prey nor double ants), belonging to no 
 fewer than three classes of the phylum Arthropoda. It is evident 
 that the predaceous foes of Stauropus, had they only the imaginative 
 powers of its human observers, could have a banquet of many diverse 
 courses, each of which would be merely Stauropus in disguise. 
 
 Poulton says: "I should not, however, have ventured to speak 
 so plainly of the meaning of the various details in the defensive 
 attitude of the larva if I had not been able to rely upon the best 
 support attainable — the support yielded by direct experiment."^' 
 He would have been approximately correct if he had said the poorest 
 support attainable, but let us see what the support is. A marmoset 
 ^nd a lizard were offered one Stauropus larva each; they showed 
 caution in attack, but each ate the larva. Rather a slender basis, one 
 would say, for four pages of argument on the special defence of the 
 creature. 
 
 The continuation of the argument — defence against insect enemies 
 — is even more far-fetched. When irritated the larva displays 
 black patches on certain segments, and Poulton thinks "it is clear 
 that the black marks exposed by the larva are calculated to suggest 
 to the approaching enemy [parasite] that the individual [larva] in 
 question is already occupied." Super-parasitism is too common an 
 occurrence to warrant the belief that parasitic insects are warned 
 away by any visible signs of preoccupation. 
 
 The tussock moths {Orgyia antiqua and 0. pudibunda) are supposed 
 to be protected by the fine hairs of the tussocks which come out 
 easily in immense numbers. Poulton says: "This interpretation 
 is entirely due to experiment. A larva of 0. antiqua was introduced 
 into a lizard's cage and, when attacked, instantly assumed the defen- 
 sive attitude. An unwary lizard seized the apparently feasible part 
 of the larva: most of the tussock came out in its mouth, and the 
 lizard seemed greatly troubled by the fine hairs and did not touch 
 the larva again" (p. 590). An Orgyia pudibunda larva was killed 
 but not eaten by another lizard. 
 
 « The BtUterflies and Moths of Europe, 1903, p. 46. 
 « Trans. Ent. Sac. Land., 1888, p. 585. 
 
304 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 The results of most of the earlier experiments with various animals 
 are brought together by Prof. E. B. Poulton in a paper entitled, 
 "The Experimental Proof of the Protective ^"alue of Color and 
 Markings in Insects with Reference to their Vertebrate Enemies" 
 (Proc. Zool. Soc. Lond., 1887, pp. 191-274). It is convenient to 
 review these experiments collectively. Those covered by Poulton 
 are as follows : 
 
 Butler, A. G. Remarks upon certain Caterpillars, etc., which are 
 Unpalatable to their Enemies. Trans. Ent. Soc. Lond., 1869, 
 pp. 27-29. 
 Animals experimented with were Lacerta viridis, frogs, and spiders. 
 In part previously published in Ent. Monthly Mag., 5, 1868, 
 pp. 131, 132. 
 Poulton, E. B. Diary of observations during 1886. P Z. S. 
 Land., 1887, pp. 269-274. 
 Using Lacerta muralis, L. viridis, and Hyla arborea. The earlier 
 sketch of some of the results of these experiments is: "Some 
 experiments upon the protection of insects from their enemies 
 by means of an unpleasant taste or smell. " Rep. British A. A. 
 S., 1886 (1887), pp. 694, 695. 
 Weir, J. Jjinner. On Insects and Insectivorous Birds, and especially 
 on the relation between the Color and the Edibility of Lepi- 
 doptera and their Larvae. Trans. Ent. Soc. Land., 1869, pp. 
 21-26. 
 The birds used were seven species of finches, one weaver bird, 
 one muscicapid, one pipit, and one thrush. Few of them are 
 highly insectivorous species. 
 Weir, J. J. Further Observations on the relation between Color and 
 the Edibility of Lepidoptera and their Larvae. Trans. Ent.. 
 Soc. Land., 1870, pp. 337-339. 
 Birds used as in his previous experiments. 
 Weir, J. J. Diary of observations during 1886. Proc. Zool. Soc. 
 Lond., 1887, pp. 268, 269. 
 Experiments with lizards: Lacerta viridis, L. agilis, and Zootoca 
 vivipara. 
 Wbismann, a. Studies in the Theory of Descent. London, 1882, 
 Vol. I, pp. 328-341. 
 Principally experimeiits with Lacerta viridis. 
 
 The main burden of the first half of Poulton's paper is the searching 
 of the results of these experiments for support of " Wallace's original 
 suggestion 'that brilliant or conspicuous larvae would be found to 
 be refused by their enemies'" (p. 196). In the preliminary sketch^* 
 
 > Rep. British A. A. S., 1886 (1887), p. 694. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 305 
 
 of his own experiments Poulton says: "Wallace had predicted that 
 brilliantly colored and conspicuous insects would be refused by the 
 ordinary vertebrate enemies of their class." This statement, which 
 was a "bull" from the very beginning (inasmuch as we cannot 
 reckon as enemies of an insect those animals that refuse to eat it) 
 is wrongly stated by Poulton in both of the above cases. Wallace's 
 original suggestion, as reported in Proc. Ent. Soc. London, 1867, 
 p. Ixxxi, is that, "as a rule, the brilliantly colored larvae were those 
 which were distasteful to birds." Poulton further twists this in his 
 table headings to a suggestion "that brilliant and conspicuous 
 larvse would be refused by some at least of their enemies," a much 
 later modification of Wallace's statement. 
 
 However, this later claim would be admitted without argument 
 did we accept Poulton's usage of the term enemies as including 
 practically all insectivorous animals. Even if there were no other 
 factors involved, the relative sizes of the larvae and of various insec- 
 tivorous animals in themselves establish limits to the number of 
 predators upon a certain form; thus numerous large larvae will be 
 free from attacks of all but a small proportion of insectivorous foes. 
 Very small larvae, on the other hand, will be overlooked by many 
 predators. That is, considerations entirely aside from coloration 
 will limit the number of enemies of any given form. It is evident 
 that all vertebrates cannot be enemies of the same insect; enemies 
 and prey form indefinite groups that intermesh in a multitude of 
 combinations. Consequently, an insect cannot be said to be pro- 
 tected, because certain vertebrates more or less ignore it, when they 
 perhaps have no opportunity and certainly in many cases no necessity 
 for feeding on it in the wild state. 
 
 Poulton first tabulates the results of experiments with eighteen 
 species of "undoubtedly conspicuous larvae," and concludes: "The 
 first and tbvious result of the first table is, with only one entirely 
 antagonistic exception, the most complete demonstration of the 
 truth of Wallace's suggestion that a highly conspicuous appearance 
 would be found to be accompanied by some unpleasant attribute" 
 (p. 205). Upon close inspection of this table, we find there are two, 
 not one, species that are not shown to be distasteful to any animal, 
 namely, Deilephila euphorbice and Lasiocampa pini; eight not refused 
 by anything are included because they were disregarded by birds. 
 The writer has explained above why disregard cannot be accepted 
 as a test at all. The inadvisability of so doing is shown by the fact 
 that at least three of these eight species of larvae, namely, Orgyia 
 
306 PROCEEDINGS OF THE ACADEMY OF [Juue, 
 
 antiqua, Cucullia verhasci,^^ and Hybernia defoliaria, are known to be 
 eaten by British wild birds. The other eight species included in the 
 table comprise three others disregarded by caged birds, two of which 
 are also eaten by British wild birds. These eight, however, were 
 used chiefly in experiments with lizards, and four of them were 
 accepted as well as refused upon trial. Of the four only refused, 
 one (Porthesia aurijlua) was eaten by hungry lizards in Poulton's 
 experiments of 1887.'^ Another, Pieris brassicce, was eaten more 
 often than refused in Pocock's experiments (reviewed later), and a 
 third which was rejected by frogs and lizards is known to be eaten 
 by nestlings of Parus major. 
 
 Exception may be taken to remarks about some of the species 
 listed in this table. For instance, Deilephila euphorbioe was eaten 
 by a captive lizard, and Newman says, "sea-gulls and terns devour 
 them in numbers." We may add to the list of enemies the malise- 
 bussard, on the authority of Schuster.^ Poulton's comment on 
 this larvae is: "The correlation of a startling appearance with some 
 unpleasant attribute must probably have existed once if not now. 
 Have we a case in which hunger or opportunity have caused the 
 enemies to neglect the latter and therefore to benefit by the former? " 
 (p. 199). We cannot so conclude, unless we admit also that similar 
 warning coloration (D. euphorbioe is "black, red, and yellow or 
 white") would lose its meaning (admitting for the purposes of argu- 
 ment that it has a meaning) to the same enemies in all other cases. 
 
 It is of interest to note that Hybernia defoliaria, included in this 
 table because disregarded by captive birds, was found in the stomachs 
 ■of three species of British birds by Robert Newstead." Schuster 
 (I. c.) records many species of birds as enemies of this larva as well 
 as of H. brumata. 
 
 Table II includes four larvse which only become conspicuous when 
 approached and detected; one is not shown to be unpalatable to 
 anything, one was both eaten and refused by lizards, and another 
 was eaten by at least two species of birds and avoided without trial 
 by two or more other species. The fourth species was refused by 
 lizards and poultry, but eaten by nestling great tits. 
 
 One of the larvae listed in this table has been made the basis of some 
 
 "See particularly the note, "Do birds eat the larvse of Cucullia?" by H. 
 D'Orville, Entomologists' Monthly Mag., VI, June, 1869, p. 16. 
 " Rep. BrUish A. A. S., 1887 (1888), p. 764. 
 M Ent. Bl. Numberg, 5, Nr. 7, July 15, 1909. 
 "Suppl. Jour. Bd. Agr. Land., XV, No. 9, December, 1908. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 307 
 
 of the most far-fetched theorizing imaginable. Chcerocampa elpenor 
 is its name; "When approached the anterior part of the body is 
 distended and resembles a serpent-like head (of the cobra type)" 
 (p. 206). In Weismann's experiments, "A tame jay ate the larva at 
 once; sparrows and chaffinches (wild) were frightened by it, and 
 would not come near a seed trough in which it was placed; fowls 
 were evidently frightened, but in the end cautiously attacked it, 
 when it was soon eaten." Lady ^'erney notes that small birds 
 "would not come near a tray with crumbs on it on which the larva 
 had been placed" (p. 206). The larva of Chcerommpa is a large one 
 (the ocellated spots are present only in last stage; if so useful, why 
 is this the case?) and its size alone is" sufficient to explain the actions 
 of the small birds. In the case of the sparrows at least, almost any 
 strange object of the same size might cause the same reaction. 
 Anything new about their regular haunts is viewed with suspicion. 
 
 In regard to the Cobra-like appearance of Chcerocampa, Poulton 
 says: "It is likely that the terrifying appearance of our own larvae 
 probably first arose in the tropics, where the imitated cause of alarm 
 to the enemies of the larvae is real and obvious. And it is probable 
 that the success of the same method in countries where the reptilian 
 fauna cannot be said to constitute a source of alarm is due to the 
 inherited memories of a tropical life which live on, as that instinctive 
 fear of anything snake-like which is so commonly exhibited by the 
 higher land vertebrates, including ourselves" (p. 204). 
 
 What a characteristic piece of selectionist reasoning(?) ; at least 
 four very debatable biological propositions, namely, the tropical 
 origin of the European fauna, its origin in a part of the tropics having 
 cobras, and instinctive fear in man and other vertebrates, are prac- 
 tically taken as established facts. Aside from these assumptions, 
 the argument is very amusing also when contrasted with that insisted 
 upon by selectionists, in a hundred places, that birds have no instinct- 
 ive knowledge of what is suitable for food, but must learn by experi- 
 ence. If an instinct of cobra fear is present in birds whose remote 
 ancestors may possibly have seen cobras, it would seem that instinct 
 about such an every-day matter as food were not a point to strain, at. 
 However, it is obvious that both arguments cannot well be sup- 
 ported by any but the exceedingly versatile. 
 
 Table III includes seven "not inconspicuous larvae which are not 
 nocturnal and which do not conceal themselves." Two are not 
 shown to be unpalatable to anything and four are included on the 
 basis of disregard by birds or lizards, at least two of which are eaten 
 
308 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 by wild birds. While the remaining one was refused by three species 
 of birds and disregarded by others, it also is eaten by wild birds. 
 
 Table IV presents the results for "bright-colored or conspicuous 
 insects other than larvae." It deals with fifteen forms, four of which 
 are not shown to be distasteful to any animal ; two are included on 
 the basis of disregard only, and four were accepted and refused by 
 the same class of enemies. One of the remaining five, i.e., Anthro- 
 cera filipendula, imago, refused by lizards in these experiments, was 
 eaten by lizards in the 1887 experiments (I.e.). Concerning another 
 insect of this group, namely, Abraxas grossulariata, Poult on notes 
 (p. 220) his opinion that Butler's record of frogs eating the moth 
 must be a mistake. It is noteworthy. However, that he uses others 
 of Butler's records without question. Butler later affirmed the 
 correctness of his note, and showed that the same insect is taken by 
 some birds. 
 
 Tables 1-IY deal with forty-four insects, nine, or 17 per cent., of 
 which are not shown to be distasteful to any animal; another nine 
 were both accepted and rejected bj- the same classof enemies. Eighteen 
 were either disregarded or rejected by birds, and at least nine of 
 these -are known to be eaten by wild British birds. We have pointed 
 out above inconsistencies of some of the other cases with other 
 experiments. In fact, as may be seen on p. 313, in seven out of eight 
 possible direct comparisons of these experiments with those of 
 Pocock, the only other extensive series using British insects, the 
 • results are inconsistent. 
 
 Eight of the insects of the distasteful groups in these tabulations 
 were fed to hungry lizards, in experiments performed by Poulton 
 in 1887,*^ and all were accepted. The behavior of the lizards in the 
 cases reported contrasts strongly with that shown in the previous 
 experiments as shown in the following table of comparisons. 
 
 Birds: , 1886. 1887. 
 
 Orgyia antiqua,\a,Tva, D A 
 
 Vanessa urticcB, pupa R A 
 
 Lizards : 
 
 Euchelia jacobce, 1 A2 R4 D A 
 
 Pygoera bucephala, 1 A3 R3 D A 
 
 Porthesia auriflua, 1 R2 A 
 
 Anthrocera filipendula, a,d R A 
 
 Abraxas grossulariata, 1 A 1 R 7 D 2 A 
 
 " " Further Experiments upon the Protective Value of Color and Markings 
 in Insects," Rep. British A. A.S., 1887 (1888), pp. 763-76,5. These experiments, 
 dealing with frogs, lizards, and a marmoset, and including a few of A. G. Butler's 
 notes on birds, are not reported in full. Hence they are not reviewed. 
 
1912.] natural sciences of philadelphia. 309 
 
 Progs : 
 
 Crcesus septentrionalis, 1 A A 
 
 Thus it appears that these experiments are inconsistent among 
 themselves, and from the fact that at least fifteen of the forty-four 
 insects alleged to have been proved distasteful by the experiments 
 are known to be eaten by wild British birds, we are justified in 
 suspecting that the experimental results do not accurately indicate 
 behavior under natural conditions. 
 
 A fifth table by Poulton comprises the results of experiments with 
 insects which are protectively colored or which evade their enemies 
 by other means. On the theory, therefore, all of these insects should 
 be eaten freely by insectivorous animals. However, in sixteen out 
 of sixty-eight cases, more than 23 per cent.,*' the "evidences of 
 distaste" are fully as strong as in the majority of the cases in the 
 first four tables. Taking this fact in connection with that previously 
 adduced, to the effect that 17 per cent, of the "protected" insects 
 were not shown to be "unpalatable" to anything, and the additional 
 fact that sixteen out of the forty-four, or 36 per cent., were included 
 on the basis of disregard (some of them being eaten also), it is quite 
 clear, that the insects, etc., were sorted out into the various tables, 
 in accordance with the requirements of the theory, experimental 
 evidence to the contrary notwithstanding. If authors are not 
 consistent in the interpretation of the results of experiments, how 
 can they expect others to have faith in them? 
 
 Poulton tries to explain the refusals of certain "protectively 
 colored" imago Lepidoptera; for instance, with regard to Vanessa 
 urticoB he says (p. 246) : "I have no doubt that the refusal of some 
 frogs was due to scales only," and of Pieris brassicce, "eaten readily 
 by all lizards, but not much relished, I believe, because of the mechani- 
 cal difficulty of the scales and wings and not from being actually 
 unpalatable." We cannot accept these explanations (which no 
 doubt are true) without asking that they be made to cover the 
 refusals of all adult Lepidoptera (of proper size for the animal experi- 
 mented with). This would affect four species of Table IV and seven 
 in Table V. We find Poulton later regarding P. brassicce as intrinsi- 
 cally unpalatable (P. Z. S., 1911, pp. 864, 865). 
 
 The inconspicuous larvse of Mania typica were tasted and rejected 
 by Lacerta muralis, Poulton's comment is: "At first sight a most 
 
 **This does not take iato account earthworms and fly larvse, which were 
 rejected as well as refused, although the fact is not stated m the table. 
 
310 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 startling difficulty, yet it is evident from the behavior of the lizards 
 that they 'fully expected the larva to be palatable, in itself a strong 
 confirmation of the suggestion that nearly all such larvae are palat- 
 able" (p. 243). This is another argument that cannot be accepted 
 unless it is also applied to the rejections of conspicuous larvse upon 
 trial, and this latter evidence rather than disregard is all the experi- 
 ments yield that is worth any consideration. In fact, if impartially 
 applied, this argument would do away with the experimental evidence 
 of the efficacy of warning colors in all cases in which the insects were 
 tasted before refusal. In other words, it would be just as fair to 
 presume that these conspicuous larvae also .were expected to be 
 palatable, "in itself a strong confirmation of the suggestion that 
 nearly all such larvae are palatable." 
 
 The experiments performed by R. I. Pocock, Superintendent of 
 the London Zoological Gardens, like those just reviewed, cover a 
 variety of vertebrate orders. A far larger number of species both 
 of predators and prey were used than in any other experiments yet 
 recorded. The captive animals included twenty-six species of 
 mammals, ninety-six of birds, and seven of lizards, of which only 
 six birds and one lizard occur naturally in England. The insects 
 used were, of course, chiefly native. Even if we believe that experi- 
 mental results have any value as indicating natural behavior, we 
 can only conclude that the conditions of these experiments invalidate 
 the findings, for of what possible value can it be to know the likes 
 and dislikes of exotic animals for British insects? 
 
 The account of these experiments is in P. Z. S. Lond., 1911, pp. 
 809-864. Mr. Pocock thinks his experimental results have "an 
 important bearing upon the criticism sometimes advanced against 
 the theory of warning coloration and mimicry as applied to butter- 
 flies, namely, that birds under natural conditions- are seldom seen to 
 eat these insects. Hence it has been inferred that birds cannot be 
 reckoned as serious enemies of butterflies. Whatever may be the 
 explanation of the circumstance," Pocock says, "I am tolerably 
 sure, from the behavior of the two classes of animals when pitted 
 against one another that the inference drawn therefrom is erroneous. 
 The insectivorous birds in our aviaries seemed to know at once what 
 the butterflies were; they were on the alert the moment one was 
 liberated and pursued it with determination and precision, following 
 its every turn and twist, and either catching it upon the wing or 
 pouncing upon it after settling. It is true that this predatory 
 deftness may have been acquired in relation to the chase of insects 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 311 
 
 other than Lepidoptera; but unless the birds recognized butterflies 
 in general — a group which cannot be mistaken for other insects — 
 as part of their natural prey, it is difficult to understand their eager 
 excitement at the sight of those I offered them" (p. 811).^' 
 
 Before quoting further, let us look into this argument a little: 
 it is characteristic of the selectionist style. He is very charitable 
 in admitting that predatory deftness may have been acquired in 
 chasing other insects than butterflies. A little reflection will con- 
 vince anyone, be he ignorant or not concerning the important con- 
 stitutents of bird food, that butterflies even if eaten, can furnish 
 but a small percentage of bird food, namely, an amount proportional 
 to their numbers among diurnal insects as a whole. Hence a 
 correspondingly small amount of training in predatory deftness can 
 possibly have been acquired from capturing them. Pocock finds 
 it difficult to understand the eager excitement of the birds at the sight 
 of Lepidoptera, unless they recognized them as such; this after 
 telling us on the preceding page of "the exceeding keenness of the 
 birds for the insects brought to them. This was no doubt due in a 
 measure to our inability in the Gardens to feed the birds on living 
 insects other than mealworms." 
 
 Caged canaries, sometimes become frantically excited when a 
 grasshopper or other insect is held up to the bars of their cage — 
 they may never have seen an insect in their life before, they only 
 know there is something they want. Pocock's parenthetical expres- 
 sion concerning Lepidoptera — " a group which cannot be mistaken 
 for other insects" — directly opposes many arguments by selec- 
 tionists relative to the resemblances of Sesiidae to Hymenoptera; 
 but any argument to establish the present point without reference 
 to its bearing on other phases of the theory is a long-standing rule 
 among selectionists. Continuing his argument, Pocock says: 
 "Again, unless the species of butterflies used for the experiments are, 
 or were in the past, habitually preyed upon by birds," whence comes 
 the extraordinary skill the liberated specimens .... displayed in 
 dodging the swoop of birds in midair? Having repeatedly seen the 
 aim of the pursuing bird baffled by the evasive twist of the butterfly, 
 I cannot doubt that the insect's behavior was prompted by the 
 instinct to escape an habitual enemy of its species, of the same class, 
 and with the same predatory methods" (p. 811). 
 
 " It is worth pointing out that the disciple is here arguing directly against 
 one of the cardinal teachings of the master, as Poulton iterates and reiterates, 
 "acceptance is not proof of palatability" {Trans. Eni. Soc. Lond., 1902, pp 436 
 317, 348, and 389). 
 
312 PROCEEDINGS OF THE ACADEMY OF [JuQC, 
 
 Suppose a different experiment were performed: let some muskrats 
 (Fiber) be put into an aquarium with some sea-lions; would their 
 -efforts to escape indicate previous experience in evading enemies of 
 -the same class? Not at all, it would indicate merely adaptation to 
 expert progression in the same medium. The relations of birds 
 (in general) to bats and of birds (again in general) to dragonflies are 
 instances illustrating the same fact, but which are due to no general 
 predatory relation between the groups. The wonderful powers of 
 flight of many of the Syrphidse are strictly comparable to that of 
 butterflies, as the adults feed only at flowers and have no need of 
 expert flight for predatory purposes; also they do not need it so 
 highly developed for defense, for rather a small proportion of birds 
 are capable of catching insects so expert on the wing. The extreme 
 rapidity and dexterity of flight of humming-birds has no possible 
 relation to their prey, nor need it have been developed to its present 
 perfection to evade species that might be inclined to prey upon 
 hummers. When Mr. Pocock arrives at the true reason for the 
 extraordinary powers of flight of humming-birds, he will undoubtedly 
 be less insistent upon the predator-evasion theory as an explanation 
 of the tortuous flight of butterflies.^ 
 
 Pocock further says: "Those who hold on the negative evidence 
 above stated, that birds are not to be reckoned as serious enemies 
 of butterflies, must be called upon to supply some explanation other 
 iihan that above proposed of the marked reactions between these 
 two classes of animals when brought into contact with one another, 
 and to show reason why what takes place in the aviary may not be 
 regarded as indicative of similar occurrences in nature" (p. 812). 
 
 The pertinent retort to this statement is that it is the selectionists 
 who first claimed and who still claim that birds are important enemies 
 of butterflies, and it is up to them to produce real evidence in favor 
 of their contention. So far they have brought forward little except 
 results of experiments. Pocock's own results, namely, the consump- 
 tion of large numbers of British insects by exotic animals, should 
 have convinced him that what takes place in the aviary may not 
 necessarily be regarded as indicative of similar occurrences in nature. 
 The point may be further illustrated by the following. Suppose a 
 
 ^ It is a matter of common observation that butterflies constantly exercise 
 their powers of flight by playing with other butterflies even of different species. 
 They often dart at faUing leaves, flying bits of paper, and even birds. The 
 -writer saw (March 27, 1912, Plummer's Id., Md.) a Vanessa antiopa dash at 
 and come within a few inches of a phoebe {Sayornis), that had just perched after 
 ■one of its customary quick sallies at insect prey. The bird, a highly insectiv- 
 ■orous species, paid no attention to the butterfly. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 313 
 
 man has fired his last shot ineffectually at a charging tiger or rhinoce- 
 ros; he naturally shows a "marked reaction" by taking to his heels, 
 not because he or his ancestors have had similar experiences, but 
 because he can run. So with the butterflies, skilful efforts to escape 
 do not necessarily indicate previous racial experience of the same 
 nature. 
 
 The principal failing of the selectionists always has been a vast 
 ignorance of what wild birds really eat. They have made very 
 little effort to acquire such knowledge, and their speculations through- 
 out show the lack of it. Practically the only large body of authentic 
 information on the natural food habits of birds is contained in the 
 records of the United States Biological Survey. They comprise 
 detailed identifications of the contents of more than 48,000 bird 
 stomachs representing all families of birds and collected in hundreds 
 of localities in the United States at all seasons. The United States 
 has a goodly representation of butterflies, yet only five of these 
 48,000 stomachs contained remains of Rhopalocera. It is hoped 
 this will be more satisfactory to the selectionists than the "negative 
 evidence" they are accustomed to cite with contempt. 
 
 The extreme artificiality of Pocock's experiments and the inappli- 
 cability of the results to the natural relations of British birds and 
 insects are so evident that it is not worth while to comment on the 
 details. A few comparisons of the results with those of experiments 
 recorded by Poulton are of interest as shomng the inconsistency, 
 inter se, of experiments. It has not been possible to collect a large 
 number of such comparisons because Poulton's experiments were 
 chiefly with lizards and few with birds, while the opposite is true of 
 Pocock's. The varjdng stages in which the insects were presented 
 also tend to limit comparisons. The table includes all possible 
 direct comparisons and onlj- one pair in eight shows real correlation. 
 
 Birds : Poulton. Pocock. 
 
 Vanessa urticce, \a.TVSi D A7R4D1 
 
 " pupa R A2R2D1 
 
 Clisiocampa neustria, larva D A 1 
 
 Euchelia jacoboe, ad A A 1 R4 
 
 Cosmotricha potatoria, larva D A 1 R 4 D 1 
 
 Anthrocera filipendula, ad A R 4 
 
 LiZARDg : 
 
 Apis mellifera, worker A R 3 
 
 Fieris rupee, ad A 20 A 2 
 
 Notes on Pocock's experiments, by Prof. E. B. Poulton, are given 
 
314 PROCEEDINGS OF THE ACADEMY OF fJune^ 
 
 on pp. 864-868, and show his customary facility in drawing conclu- 
 sions satisfactory to himself from the most refractory evidence. 
 For instance, he says: "The experiments on the Pierince support 
 the conclusion that the perfection of the under surface procryptic 
 resemblance affords a true criterion of the degree of palatability. 
 Pieris brassicce, with its conspicuous gregarious larva and imago 
 larger and less cryptically colored than the other three species .... 
 was distinctly the least palatable" (pp. 864, 865). The records of 
 acceptances and rejections of the three species of Pieris are as follows : 
 Pieris rupee, A 10 R 2, a proportion of 5 to 1 ; Pieris brassicce, A 40 
 R 33, 1^ to 1; and Pieris napi, A 8 R 8, 1 to 1; P- napi thus being 
 the least favored in the experiments. These figures are reproduced' 
 not because the point as to relative palatability is of any importance, 
 especially as an indication of natural preferences, but merely to sliow^ 
 that the experimental results are not accurately judged by those 
 most interested in them. It is of interest to compare Pocock's 
 results with Pieris brassicce with those obtained by Dr. G. Rorig in 
 Germany. Pocock records the pupae of this insect as A 1 R 8 D 3, 
 and the adult as A 16 R 7 D 1, while Dr. Rorig says:^' The pupae 
 of the Kohlweisling "were eagerly torn open by all the titmice, " and 
 the adults "were always freely eaten by all the birds which I have 
 tested." 
 
 Poulton bases considerable speculation upon the unpalatability 
 of Araschnia levana, the record for which in the experiments is A 20 
 R 10. He follows this with a page of theorizing on the probable 
 mimicry of Melitcea by Hesperia, and says that the experimental 
 "results as a whole leave little doubt that Melitcea is distasteful to 
 many birds, and that it does actually possess the qualtities which 
 would render it an advantageous model for the Hesperiidae" (p. 867). 
 As a matter of fact, the experiments with birds and Melitcea de- 
 scribed on pp. 826 and 827 show that it was finally refused by only 
 one bird; it was eaten by thirteen species, seven species of which 
 took nine specimens without hesitation. Two birds which dropped 
 the first specimen offered them later took one and two, respectively, 
 including those dropped. The final record for birds is A 18 R 1. 
 If this is considered proof that Melitcea is advantageous as a model, 
 the demands of the theory are most modest. 
 
 On p. 867 he also refers to Melanargia as a highly distasteful genus ; 
 its record in the experiments is A 14 R 4. Another extract from 
 
 ^ Arb. Biol. Abt. f. Land. Forslunrls. K. Gesundheit.samte, 4, 1903, p. 47. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 315 
 
 Poulton's remarks is, "Although so many insectivorous animals in 
 confinement disregarded the special defence of Formica rufa, there 
 can be little doubt that such defence is very effective in the wild 
 state. It is impossible on any other hypothesis to account for the 
 conditions under which the species exists, swarming in vast numbers 
 in restricted areas and an easy prey to any enemy that would dare 
 to attack" (p. 868). Here we have a case where experimental 
 results are not in accord with the theory, and it is evident that it is 
 so much the worse for the experiments. Where the evidence is of 
 a supporting nature, experiments are extolled to the skies. Pocock, 
 who tries to stick consistently to the experimental results which for 
 Formica rufa were an indefinite number of acceptances and no 
 rejections by a monkey, A 13 R by birds, and two refusals by 
 lizards, says: "The unavoidable conclusion that these insects are 
 palatable is rather surprising in view of the frequency with which 
 ants are mimicked in the tropics " (p. 849). In deeming it impossible 
 for the ants to live as at present unless specially defended, Poulton 
 takes the struggle for existence too seriously. In fact, he seems to 
 think all gregarious insects must be specially protected, thus over- 
 looking periodical cicadas, migratory locusts, canker-worms, army- 
 worms, etc., which usually occur in large numbers and are eagerly 
 attacked by a great variety of insectivorous foes. 
 
 Poulton further remarks: "It was also apparent in many of the 
 experiments that the unpalatability of conspicuous Lepidoptera 
 was .... far more obvious to the birds than the mammals. In 
 view of the part which birds are believed to play in the production 
 of mimetic resemblances, it is obvious that this inference may be 
 highly significant" (p. 868). The writer has tabulated the accept- 
 ances and rejections for mammals and birds, including only those 
 Lepidoptera which were refused by some species, and the result is 
 mammals A 19 R 10, or about 34 per cent, refused, and birds A 112 
 R 80, or about 41 per cent, refused. Hardly enough difference to 
 warrant the comment quoted. 
 
 A very interesting series of experiments with frogs, lizards, birds, 
 and mammals upon a gDod variety of insects and other inverte- 
 brates as subjects were performed by Beddard and Finn at the 
 London Zoological Gardens and recorded by the former in his volume 
 entitled Animal Coloration (1892, pp. 149-166). Mr. Beddard's 
 principal conclusions are as follows: "It is quite clear from these 
 experiments that insects which exhibit warning colors are by no 
 means always exempt from attack. The opinions of insect-eating. 
 
316 PBOCEEDINGS OF THE ACADEMY OF [June, 
 
 mammals, birds, and reptiles appear to vary as to the edibility of 
 
 this or that insect But these experiments do show that 
 
 very generally, though not always, a disagreeable taste is associated 
 with a conspicuous and varied coloration. On the other hand, 
 precisely the same deductions can be drawn by watching the behavior 
 of animals when offered inconspicuously colored insects" (p. 155). 
 
 As in previous cases, we will give the direct comparisons that can 
 be made between these and other sets of experiments. Four out of 
 seven contrasted pairs are contradictory. 
 
 Poulton's 
 
 Tables, Animals 
 
 1887. Beddard. tested. 
 
 Armadillo vulgaris A A 2 Lacerta viridis. 
 
 Lithobius forficatus A R , " " 
 
 Pieris brassicoe, 1 R A 5 R 2 D 1 Lizards. 
 
 Abraxas grossulariata, 1 .. A 1R7D2R1D2 " 
 
 Vespa vulgaris D A 2 " 
 
 Euchelia jacobw, 1 A 2 R 4 A 1 R 3 D 1 
 
 Pocock, 1911. Beddard. 
 Pieris brassicoe, 1 A 20 R 18 A 2 D 1 Birds. 
 
 Beddard justly remarks (p. 166): "Xone of these experiments 
 are thoroughly satisfactory; it is so difficult to interpret them, and 
 they are often contradictory, for a bird will eat one day what it has 
 refused before. The experiments that have been made are like most 
 other statistics — they may be made to prove anything. A careful 
 series of observations upon the contents of the stomachs of wild 
 birds would be the nearest approach to a satisfactory solution of the 
 difficulty; but there are obvious objections to this mode of inves- 
 tigation." 
 
 Fortunately, this objectionable method has been pursued to some 
 extent in England, i.e., by Newstead, and to a slight degree the 
 work serves as a check on experiments with British birds and insects. 
 Beddard gave an earwig to a green woodpecker, which made a great 
 deal of fuss over it, but ended by swallowing it; Newstead found 
 these insects in two stomachs of green woodpeckers, one of which 
 contained 23.^ 
 
 Merely for the sake of completeness the very brief notes upon 
 experiments by Thomas Belt'' may be given here: 
 
 A tame white-faced monkey always killed but did not eat Heliconii 
 (pp. 316, 317). 
 
 «> Suppl. Joum. Bd. Agr. [Lond.], XV, 1908, p. 64. 
 " The Naturalist in Nicaragua, 1888. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 317 
 
 Lampyridse were "invariably rejected by the monkey, and my 
 fowls would not touch them") p. 317). 
 
 A red and blue frog was not touched by fowls and ducks, but one 
 of the latter tricked into taking one rejected it (p. 321). 
 
 We may add also those of Haase:'^ 
 
 Erithacus rubecula had to be starved into eating Zygcena trifolii 
 (p. 20). 
 
 Fowls always rejected Danais chrysippus, but eagerly ate Papilio 
 pammon. Captive monkeys also rejected Danais (p. 23). 
 
 Attempts to feed species of Danais, Pharmacophagus, Delias, and 
 Euschema to tame chickens were unsuccessful (p. 99). 
 
 Danais septentrionalis was rejected by caged lizards (Calotes 
 mystaceus) (p. 99). 
 
 The following, appended in bibliographic form, arc self-explana- 
 tory: 
 
 DoNiSTHORPE, H. St. J. K. Cases of Protective Resemblance, 
 Mimicry, etc., in the British Coleoptera. Trans. Ent. Soc. 
 Land., 1901. 
 
 Three species of lizards were found to reject the Telephorid, 
 Psilothrix nobilis (p. 362). 
 
 Specimens of Mdasoma populi "were pecked at, but finally refused, 
 though killed, by Shama, Pied Mynah, Laughing Jackass and 
 Brambling. The Drongo and Graculipica nigrirostris ate 
 several." The author says: "It appears to me that their 
 refusal by so many insect-eaters in confinement conclusively 
 proves their distastef ulness " (p. 368). It proves nothing of 
 the sort. Moreover, " distastef ulness " of the kind here claimed 
 seems to be no protection at all. Specimens of Cassida equestris 
 were eaten by all the birds to which they were offered (p. 369). 
 Shelford, R. Observations on some Mimetic Insects and Spiders 
 from Borneo. Proc. Zool. Soc. Land., 1902, pp. 230-284, 
 pis. 19-23. 
 
 Two tame monkeys (Macacus cynomolgus) manifested disgust 
 after tasting specimens of the Reduviid {Eulyes amaena), but 
 ate its mimic, the. mantis {Hymenopus bicornis) (p. 232). The 
 writer says he has proved the distastefulness of Lycidse, by 
 repeated trials with various small mammals and birds (p. 244). 
 "AH the Lycidse are strongly distasteful .... A strong vitality 
 is correlated with this distastefulness: I have seen a Lycid 
 beetle walk away apparently uninjured after it had been well 
 pecked by two or three fowls" (p. 267). 
 TiTCHENER, E. B., and F. Finn. Comparative Palatability of 
 Insects, etc. Nature, Vol. 42, No. 1,093, October 9, 1890, 
 pp. 571, 572. 
 
 ^ Haase, E., Researches on Mimicry, Part II, Stuttgart. 1S96. 
 
318 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 The animals used in these experiments were domestic mice, toads, 
 a mynah (Acridotheres tristis), a heron (Ardea cinerea), a prairie 
 owl, a water tortoise and a lizard. The results of the experi- 
 ments are described in detail, but no general conclusion is given. 
 At least seven of the things offered as food were both accepted 
 and refused by the same species of animal. This number 
 included the common earthworm (Lumhricus terrestris) . 
 
 TiTCHENER, E. B. Comparative Palatability. Nature, Vol. 44, 
 No. 23, October 8, 1891, p. 540. 
 Experiments with frogs, toads and ducks, supplementary to the 
 above; no general remarks. 
 
 TiTCHENER, E. B. Comparative Palatability. Nature, Vol. 45, 
 No. 3, November 19, 1891, p. 53. 
 These experiments relate to the choice of food by captive goldfish, 
 silverfish, frogs, and a spider. The details are given without 
 comment. 
 
 BIRDS. 
 
 Experiments in Europe. 
 
 Birds have been used more frequently than animals of any other 
 class to test the potency of the protective adaptations of insects and 
 other groups under experimental conditions. One of the most 
 important series of experiments was carried on chiefly as a study of 
 the origin of the process by which food is accepted or rejected by 
 birds. In this series Prof. C. Lloyd Morgan performed various 
 experiments with young chicks, pheasants, guinea-fowls, moorhens, 
 and ducks, the net result of which "is that, in the absence of parental 
 guidance, the young birds have to learn for themselves what is good 
 to eat and what is distasteful, and have no instinctive aversions."^ 
 The results of these experiments are often quoted by the selectionists, 
 and as usual in such cases with sweeping inclusions not at all intended 
 by the author. He says: "I am not, of course, prepared to say 
 
 that in no case is there such instinctive aversion Birds 
 
 like the megapodes, which are hatched out in mounds apart from 
 parental influence .... may show instinctive avoidances which 
 our well-cared-for birds do not possess. That the parent bird does 
 in most cases afford guidance is unquestionable" (pp. 43-44). 
 
 Some of the principal results that have a bearing on the value of 
 warning colors under experimental conditions are as follows: 
 
 1. Chicks tested and rejected cinnabar caterpillars (Euchelia jacobm) , 
 but ate brown loopers and larvae of the green cabbage-moth 
 (p. 42). A jay ate five cinnabar larvae, but would take no 
 more (p. 43). 
 
 " Habit and Instinct, 1896, p. 43. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 319 
 
 2. Young moorhens found the conspicuously colored burnet moths 
 
 {Zygcena filipendula) distasteful, the obnoxious part being the 
 wings, for the body from which the wings were removed was 
 eaten with apparent relish while the severed wings were rejected 
 (p. 42). 
 
 3. Lumbricus fcetidus was refused at first, as were all other earth- 
 
 worms for some time afterwards. Later all were eaten. 
 
 4. All birds tested avoided woolly-bear caterpillars (Arctia caja). 
 
 5. Jays, ducks, and moorhens ate caterpillars of the tiger moths, 
 
 Nemeophila plantaginus and Chelonia villica, while chicks, 
 pheasants, and guinea-fowls found them distasteful (p. 43). . 
 
 6. Jays ate pupae of Abraxas grossulariata (p. 43), an insect refused 
 
 by most of the captive animals to which it has been offered. 
 
 7. One jay ate a larva of Phalera bucephala, which another jay and 
 
 duckling and chicks refused (p. 43). 
 
 8. Soldier-beetles and ladybirds were avoided (p. 43). 
 
 In No. 2, intrinsic distastefulness is not shown; it is the dry, scaly 
 wings that are objectionable. Nos. 3, 5, and 7 give evidence for 
 both sides of the question, and No. 6 is inconsistent with most other 
 experiments on the same insect. 
 
 The evanescence of some associations concerning food are shown 
 by the following experiment: Bits of orange peel were offered to a 
 young chick that had learned to eat yolk of egg; they were refused, 
 as were also bits of yolk substituted immediately afterward. Sub- 
 sequently the yolk was again tested and accepted (p. 41). 
 
 Another test indicates that in some cases taste cannot be the 
 criterion upon which choice is made. "While small worms are 
 picked up with avidity, large worms are left alone by quite, young 
 birds and often evoke the alarm note. None of the chicks on the 
 fifth day dared go near a particularly large worm. Bits of red-brown 
 worsted, somewhat resembling worms, were seized with eagerness 
 and eaten with surprising avidity so long as they were not more than 
 a couple of inches long. Of a four-inch bit the chicks were afraid, 
 until one bolder than the rest, seized it, whereupon the other chased 
 him for the prize till he escaped to a secluded corner and swallowed 
 it"(p. 50). 
 
 An unusual experience with customary food may lead to its rejec- 
 tion, as decisively as if it were "nauseous" and " warningly colored." 
 "Pheasants and partridges, when they seized a worm for the first 
 time, shook it and dashed it against the ground; one of them did so, 
 indeed, with such vigor that he shook himself over, and thereafter 
 could not for some time be induced so much as to look at a worm" 
 (p. 51). "A little pheasant which would run to my hand for wasp 
 
320 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 larvae placed upon the palm, one morning gave the alarm note, and 
 would not as usual jump upon my fingers. Four or five of the 
 grubs had stuck together so as to form a large mass of which he was 
 afraid!" 
 
 "Moorhen chicks were at first afraid of the common yellow under- 
 wing moth and of the gamma moth, though both were eaten freely 
 after I had given them dead moths" (p. 50). 
 
 "Even protective coloration is of little value if there is movement, 
 so sharp are the eyes of young birds. The caterpillar of the small 
 white butterfly {Pieris rapoe) on a nasturtium leaf, with which its 
 clear green color assimilated well, was picked off by a moorhen chick 
 the moment it moved its head. Recently hatched stick insects 
 {Diapheromera femorata) , which Prof. Poulton gave me, were snapped 
 off the lime leaves directly they moved" (p. 46). 
 
 Prof. Morgan made many tests with bees and wasps, and in 
 summing them up says: "Much .... depends on the nature 
 of initial experience. A bird that has in early days seized a bee 
 with ill effects is shy for a long time, not only of bees, but of moths, 
 large flies, and beetles, while one which is so stung at a later stage is 
 made, perhaps, a little more cautious generally, but the main effect 
 is a particularized one concerning bees or the bee-like drone fly" 
 (p. 54). 
 
 A series of experiments, of much the same nature as Morgan's, 
 but shorter, is described by L. W. Kline." Chicks were tested with 
 earthworms, white boring grubs, cabbage worms, and bits of yellow 
 pine and starched muslin. "They rejected pine wood after a few 
 experiences at the age of three days, but three days later they ate 
 it again, while experience ■with muslin on the third day was lasting. 
 They were six days getting acquainted with earthworms and eight 
 days with canker [cabbage] worms" (p. 276). 
 
 An excellent article, previously referred to, "The Food of Some 
 British Birds, "^ by Robert Newstead, besides presenting the largest 
 amount of detailed information on its subject, thus far brought 
 forward, contains a short account of an experimental feeding of 
 starlings. Certain food items were placed near a nest in which 
 young were being fed. One centipede {Geophilus longicornis) and 
 one earwig {Forficula auricularia) were refused, although.each species 
 had previously been given to the nestlings by the parent birds. Only 
 
 "" Methods in Animal Psychology " [Chicks], Amer. Journ. of Psychol., V 
 1898-9, pp. 265-277. 
 
 ^Suppl. Journ. Bd. Agr. [London], XV, Xo. 9, December, 1908. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 321 
 
 one of six wood lice (Oniscus asellus) was accepted, and from five to 
 seven green cherries were refused. Both of these items are eaten by 
 adult starlings, thus rounding out to a total the contradictory 
 evidence as to choice of the four items by the same bird under natural 
 and under artificial conditions. 
 
 In 1889 and 1890, Mr. A. 'G. Butler, whose experiments with 
 lizards are included in the tables of Poulton, previously discussed ^ 
 published three articles dealing with the food preferences of captive 
 birds. These included both British and tropical birds, which were 
 kept in large aviaries. The first** of Mr. Butler's trio of papers 
 treats only the general results of six years' experimenting. He says : 
 "My experience .... has been that no insect in any stage was 
 ever refused by all the birds; what one bird refused another would 
 eat" (p. 171). In the course of this paper, Mr. Butler casually 
 remarked that for two years he had sent data on the experiments to 
 Mr. Poulton, "not even retaining a copy of my notes, but so far 
 nothing seems to have come of it; I presume, therefore, that my facts 
 have rather tended to mystify than clear the matter up" (p. 171). 
 Poulton seems to have taken deep umbrage at Butler's remarks,, 
 as he returned the notes and made a hot reply on pp. 358-360 of the 
 same volume. Butler later published his notes in full,'' and says: 
 "Few things ever astonished me more than the hostile attitude which 
 Mr. Poulton assumed with regard to that innocent paper, or the 
 cruel misconstruction which he put upon the most harmless remarks 
 made therein; that my comment touching the repeated reproduction 
 of a few comparatively unimportant observations of my own should 
 have been dislocated into a claim to the origination of Wallace's 
 
 theory is too absurd to be considered seriously I still 
 
 insist that, so long as a few desultory observations are incessantly 
 forced into a front place, it is an evidence of how little has hitherto 
 been done upon which to establish the truth of a theory; many more 
 observers are wanted, and all their observations must be impartially 
 treated if we are to arrive at exact scientific truth. I was not aware 
 that Mr. Poulton had made a selection of 'the most interesting 
 results' of my recent experiments for publication in the Report of 
 the British Association, or I should not have said 'so far nothing 
 seems to have come of it'; nevertheless, as it is impossible for any 
 
 ^ "A few remarks respecting Insects supposed to be distasteful to Birds," Annals 
 and Mag. of Nat. Hist., Sixth Ser., Vol. IV, 1889, pp. 171-173. 
 
 " " Notes made during the summer of 1887 on the effect of offering various. 
 Insects, Larvse, and Pupse to Birds," I.e., pp. 463-473. 
 
322 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 one man to judge how far even apparently uninteresting results maj- 
 eventually tell for or against a theory — as, too, Mr. Poulton has 
 evidently forgotten some of those facts .... I think I cannot do 
 better than publish the whole of my observations in detail" (pp. 
 463, 464). 
 
 "The most interesting results" made use of by Poulton^ are 
 remarks upon only four species of insects eaten by birds, while 
 Butler's notes deal with at least forty-seven species of insects and 
 other invertebrates. Moreover, without mentioning Butler's results, 
 Poulton discusses the results of his own tests with lizards and a 
 marmoset of three other species of insects, which Butler had fed to 
 birds. Poulton gratuitously observes: "If I had no more notes 
 than those supplied by Mr. Butler, their preparation for publication 
 would be only a work of a few hours; but these notes are a very 
 small fraction of the whole."*^ The fact remains, however, that the 
 large "fraction of the whole," with unimportant exceptions remains 
 unpublished to-day. As a result of this series of experiments, 
 Butler concludes that "no insect in any stage, excepting the red- 
 tailed bUmble-bee (which, by the way, I only offered to the missel- 
 thrush), was rejected by all my birds; those insects which were 
 
 refused by certain species were eagerly devoured by others 
 
 In the second place, so far from my birds learning by experience to 
 reject with scorn that which they had proved to be unpalatable, 
 I found that in some instances they seemed to acquire a taste for 
 larvse previously refused. Birds are very intelligent, but their 
 memories are ridiculously short" (p. 473). 
 
 Butler's third paper'" enumerates tests of 17 invertebrates offered 
 to birds, with the following principal conclusion: "My experiments 
 have convinced me that the tastes of birds not only differ in individ- 
 uals of the same species, but that the same individuals in consecutive 
 years vary as to their likes and dislikes." 
 
 Unfortunately, the experiments of Butler cannot be compared 
 with those of Pocock, who also used British insects and both native 
 and exotic birds, as Butler does not record the number of times an 
 insect was refused or accepted, but only tells what species of birds 
 ate it and which did not. Probably the only coincidence of the same 
 
 « Rep. BrUish A. A. S., 1887 (1888), pp. 762, 763. 
 
 » Ann. and Mag. Nat. Hist., 1889, pp. 359, 360. 
 
 ™ " Notes made during the present year on the Acceptance or Rejection of 
 Insects by Birds," Ann. and Mag. Nat. Hist., Sixth Ser., Vol. VI, 1890, pp. 
 324-327. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 323 
 
 species of bird tested with the same stage of the sapie species of 
 insect, in the two sets of experiments, is Leiothrix with larvae of 
 Pieris brassicce. The result in each case was acceptance. 
 
 A. D. Bartlett recounts an experience in rearing young water 
 ouzels which well illustrates the fundamental difference between 
 experimental and natural conditions. He says:'' 
 
 "They had been tried with the usual food for most insect-eating 
 birds, such as scraped beef and hard-boiled eggs, ant eggs, mealworms, 
 spiders, flies, beetles, aquatic snails, shrimps, salmon spawn, and 
 many other mixtures, but all failed, until my clerk and assistant, 
 Mr. Arthur Thomson, who had taken as much interest in rearing 
 these birds as myself, hit upon the idea of scalding the -mealworms, 
 and tried it. It was soon apparent that in this condition the meal- 
 worms could be digested, while in a raw or living state they (espe- 
 cially their hard skins) would pass through the birds in a hard and 
 undigested condition. From this moment I had but little trouble. 
 The birds fed greedily upon the half-boiled mealworms, and I soon 
 found them ready to leave the nest." 
 
 Thus these birds did not thrive upon a regimen that included 
 several elements of their natural food, but did well only when the 
 staple food was partially cooked. j\Ir. Bartlett adds: 
 
 "In May, 1869, I obtained my first living water ouzel. Since 
 that time I have had a great many of these birds. Some of them 
 I reared from the nest, and I fed them upon boiled mealworms, the 
 larvae of the caddis fly and other insect food; but as soon as they 
 were able to feed themselves and took to the water, they caught and 
 fed upon very small fish, especially yoimg minnows. I found them 
 rather expensive pets, having to provide for a family of four, as they 
 caught and devoured several dozen daih', and seemed to prefer live 
 fish to all other food." 
 
 If experimental results could be taken as a guide to natural be- 
 havior, we should conclude from this testimony that water ouzels 
 feed largely on fish. It is worthy of note, therefore, that Newstead'^ 
 found no fish in the stomachs he examined. 
 
 In the account" of the experiments by Dr. G. Rorig, previously 
 referred to, it is stated that all of the following insects : 
 
 " Wild Animals in Captivity, 1899, pp. 308-310. 
 « Suppl. Journ. Bd. Agr. Land., XV, No. 9, December, 1908, p. 25. 
 " Arb. Biol. Abt. f. Land. u. Forstwirtschaft. K. Gesundheitsamte, IV, 1903, 
 Heft 1, pp. 34-50. 
 
324 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 Cnethocampa pinivora, eggs and larvae; 
 Fidonia piniaria, larvae; 
 Euprodis chrysorrhoea, larvae; 
 Clisiocampa neustria, pupae and adults; 
 Liparis salicis, pupae and adults; 
 Pier is brassicoe, pupae and adults; 
 Porthetria dispar, adults; 
 Xematus abieium, larvae; 
 Nematus salicis, larvae, 
 
 were taken eagerly by captive birds, such as titmice, redstarts, 
 kinglets, nuthatches, etc. Although the list includes hairy larvae, 
 some ■nith urticating hairs, and sawfly larvae which other experi- 
 menters state that birds usually reject. Dr. Rorig does not mention 
 any refusals. We have already quoted his notes on the accept- 
 ance of Pieris brassicce, which has been classed as distasteful. 
 Dr. Rorig's birds also ate plant-lice, Aradus cinnamomeus, Ceddomyia 
 saliciperda, Retinia buoliana, R. turionana, Phyllopertha horticola, and 
 Scolytidce. 
 
 We may note here also the experiment'^ of Dr. Gtinther in feeding 
 Meloidae to chickens. He fed the fowls from 1 to 5 grams of Can- 
 tharus daily, until a total of 28, 28, 40.5 and 80.5 g. of the material 
 was taken by four chickens respectively. One of the birds which 
 ate 28 g. showed symptoms of poisoning; the others remained 
 healthy. Significant amounts of cantharidin were recovered from 
 the bodies of these birds, and even from eggs laid by them. 
 
 Another German experiment but slightly related to the theory 
 of protective adaptations is recorded'^ by Alexander Bau. The 
 titmice, Parus major and P communis, accepted in confinement 
 eggs of Liparis monacha, Porthetria dispar, Orgyia spp., and Clisio- 
 campa neustria (p. 69). 
 
 Brief mention should be made of the following: 
 
 [DoNiSTHORPE, H.] [Experiments with Birds.] Proc. Ent. Soc. 
 Lond., 1901, p. xiii. 
 Quoted by Rev. Canon Fowler, to the effect that Clythra quadri- 
 punctata, Gonioctena rufipes, and species of Ldna were rejected 
 by several species of British and foreign birds in the London 
 Zoological Gardens. All of these beetles were eaten by a 
 racket-tailed drongo. 
 
 LoNGSTAFF, G. B. Experimental evidence as to the Palatability 
 of Butterflies. Trans. Ent. Soc. Lond., 1908, pp. 629-631. 
 
 " Tier&rztliches Zentralhl, 34, Nr. 18, June 20, 1911, S. 273-276. 
 ™ "Nutzen und Schaden durch die Vogel; Vogelschutz." In Naturgeschichte 
 der Deuischen Vogel, by C. G. Friderich, Stuttgart, 1905, pp. 60-76. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 325 
 
 The experiments were performed in Ceylon with Gracula sp., 
 and domestic fowls. "So far as these experiments teach any- 
 thing, it would appear that these mainas would eat with relish 
 Nissanga patina, Yphthima ceylonica, Atella phalanta, Ergolis 
 sp., and Lampides sp. On the other hand, Papilio aristolochice 
 and Crastia asela were distinctly distasteful. The evidence 
 as to the other species experimented with fails to convince me 
 one way or the other" (p. 631). In several of the experiments 
 the birds apparently were not hungry enough to care for anj^- 
 thing. 
 
 Experiments in Africa. 
 
 In their extensive and interesting paper on the " Bionomics of 
 South African Insects,"'* Marshall and Poulton record the results 
 ■of experiments with kestrels {Cerchneis rupicoloides and C. naumanni) 
 and a ground horn-bill {Bucorax caffer). 
 
 The experiments with the kestrels (pp. 340-345) are characterized 
 by the average small number of trials of the various insects used. 
 The writer desires to draw attention to only one point in the dis- 
 cussion of these experiments. On p. 346, Poulton says with regard 
 to some supposedly distasteful beetles which the birds had eaten: 
 "It is probable that most of the defensive fluid had been ahead}' 
 discharged in the case of the Carabids of the genera Piezia, Poly- 
 hirma,, and Graphipterus, of which the acid secretion was seen to be 
 a very positive protection when there was opportunity for its opera- 
 tion on a normal scale." The "normal scale" referred to was the 
 offering oi the beetles tail first! Marshall found these carabids in 
 the stomachs of certain wild 'birds, and in discussing this Poulton 
 says the fact is not remarkable, as "the defensive secretions may be 
 discharged and lost as the result of the attacks of an experienced 
 enemy" (p. 353). This better illustrates action on a "normal 
 scale." 
 
 The ground horn-bill experimented .upon by Marshall (pp. 347- 
 348) ate all butterflies offered it, including several of the reputed 
 "protected" forms, with the exception of two specimens of Limnas 
 (Danais) chrysippus. Poulton, therefore, remarks: " It has already 
 been pointed out that the acceptance of insects by insectivorous 
 animals in captivity is no proof of their normal likes or dislikes in a 
 
 wild state Hence the fact the Acrseas were devoured is no 
 
 evidence that they are normally eaten except in a time of unusual 
 hunger" (p. 348). Alarshall," however, says: "The bird was 
 
 '6 Trans. Ent.Soc. Lond., 1902, pp. 287-504. 
 " Trans. Eril. Soc. Lond., 1908, p. 139. 
 
326 PROCEEDINGS OF THE ACADEMY OF [JuUe, 
 
 entirely unconfined, and wandered at will searching for its food just 
 like his wild relatives on the next hillside, with only this exception, 
 if insects, etc., were scarce, the .... bird always got additional 
 food at the house. The conditions of the experiment, therefore, 
 render it highly improbable that the hornbill was eating insects 
 which it would normally reject, and its whole demeanor was quite 
 at variance with such a supposition." 
 
 It is apparent that experts may draw very different conclusions 
 from the same experimental data, a fact among many which points 
 to the conclusion that the results of stomach examination are the 
 only reliable criteria regarding bird food. 
 
 Experiments in Asia. 
 
 / 
 
 No experiments are more widely quoted than those performed by 
 Frank Finn while Deputy Superintendent of the Indian Museum, 
 Calcutta. Only those of Pocock are more extensive, and they were 
 performed under much more artificial conditions. The results of 
 Finn's experiments on birds are published in the Journal- of the 
 Asiatic Society of Bengal, as follows: 
 
 No. I. Experiments with a Babbler {Crateropus canoriis). Vol. 64, 
 1895 (1896), Pt. 2, pp. 344-356. 
 
 No. IV. Experiments with various Birds. Summary and con- 
 clusions. Vol. 66, 1897 (1898), Pt. II, pp. 613-668. 
 
 The birds used in these experiments were; 
 
 Pekin robin. Leiothrix luteus. 
 
 Common babbler Crateropus canorus. 
 
 Red-whiskered bulbul Otocompsa emeria. 
 
 Common bulbuL.... Molpastes bengalensis. 
 
 Yellow-vented bulbul Molpastes leucotis. 
 
 White-crested bulbul Pycnonotus sinensis. 
 
 Green bulbul Chloropsis sp. 
 
 White-eye Zosterops sp. 
 
 Sibia . Malacias capistrata. 
 
 Mesia Mesia argentauris. 
 
 Button quail Turnix taigoor. 
 
 Bhimraj... Dissemurus paradiseus 
 
 King-crow Dicrurus ater. 
 
 Shama ..Kittacincla macroura. 
 
 Indian starling... ...Sturnus menzbieri. 
 
 Mynah Acridotheres tristis. 
 
 Black and white hornbill Anthracoceros sp. 
 
 The Zosterops, probably because of their small size, played a very 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 327 
 
 minor part in the experiments, caged as they were with a variety 
 of larger birds; the Mesia had an exceedingly brief trial, and the 
 Anthracoceros was tested principally with dead and dry insects left 
 over from experiments with other birds. Finn remarks that of two 
 individuals of this last species, one did not care for insects at all; 
 the other on some occasions had to be coerced into eating insects 
 of the supposedly palatable kinds. 
 
 In commenting upon the significance of his experiments Finn puts 
 an emphasis on order of choice between insects, which the writer 
 does not consider justifiable. These as all other experiments are 
 to determine what will be eaten, and the fact that insects not eaten 
 in the presence of the experimenter, as a rule, were devoured before 
 his next visit or the next morning, shows the futility of drawing fine 
 distinctions as to apparent preferences. 
 
 More than 123 butterflies which were left in the cages were eaten 
 in the absence of the experimenter or by birds not specified, and 
 more than 77 per cent, of them belonged to the "nauseous" group. 
 About seventy-two butterflies remained uneaten overnight, though 
 it should be remarked that many of these were taken later the next 
 day. Of the seventy-two, about 85 per cent, belonged to the "nau- 
 seous" group, a percentage practically no different than in the case 
 of those eaten. It is worthy of note that the number of butterflies 
 left uneaten is definitely stated in every case, while those eaten are 
 often included in general terms, as "some," "several," etc. In 
 making these calculations, "some" has been reckoned as two; 
 undoubtedly it sometimes meant more. These expressions occur 
 nineteen times for the "nauseous" group among the butterflies 
 eaten in the absence of experimenter, only once for the "palatable" 
 group, and not at all in the case of butterflies left over. Hence 
 there is no doubt as implied above that the proportions of these 
 groups are about the same in the butterflies eaten as in those left 
 over. 
 
 In the following table are shown the approximate numbers of 
 acceptances and rejections upon trial of both the "nauseous" and 
 "palatable" groups of insects. Species of the former group used are 
 Acroea violce, Danais chrysippus, D. genutia, D. limniace, Delias 
 eucharis, Euplcea sp., Euproctis sp., Mylabris sp., Papilio aristo- 
 lochice, and Terias sp. The principal species of the "palatable" 
 group are Catopsilia sp., Junonia sp., Elymnias undularis, Papilio 
 demoleus, P. poliies, Huphina phryne, Hypolimnas misippus, Nephe- 
 ronia hippia, Atella phalanta, and Neptis kamarupa. 
 
."328 PROCEEDINGS OF THE ACADEMY OF [JuUC, 
 
 Xauseous. 
 
 A. R. 
 
 Crateropus canorus 111+ 36 
 
 Leiothrix luteus 52 22 
 
 Dissemurus paradiseus 30+ 22 
 
 Dicrurus ater 17 2 
 
 Kittadncla macroura 21 28 
 
 Stumus menzbieri 17 16 
 
 Chloropsis sp 6 5 
 
 Maladas capistrata 
 
 Otocompsa emeria'^ 21 7 
 
 Molpastes bengalensis''^ 15 1 
 
 Molpastes leucotis 15 
 
 Pycnonotus sinensis 7 9 
 
 Tumix taigoor 29 1 
 
 Acridotheres tristis 2 1 
 
 A. 
 
 R. 
 
 96 
 
 1 
 
 94 
 
 21 
 
 54 
 
 6 
 
 32 
 
 4 
 
 78 
 
 12 
 
 42 
 
 7 
 
 16 
 
 5 
 
 5 
 
 5 
 
 7 
 
 6 
 
 10 
 
 
 
 5 
 
 
 
 10 
 
 1 
 
 20 
 
 3 
 
 2 
 
 1 
 
 343+ 150 471 72 
 
 Finn's conclusions may be discussed in order: 
 
 1. "That there is a general appetite for butterflies among insec- 
 tivorous birds, even though they are rarely seen when wild to attack 
 them" (p. 667). 
 
 This is a thing which can never be proven by experiment. As 
 well say there is a general appetite for boiled rice, bread and milk, 
 and domestic cockroaches which were the stock foods of the birds 
 used in these experiments. Certainly, these things are no more 
 foreign to the natural dietaries of many species of birds than are 
 butterflies, and the eating of either in captivity is no proof that they 
 are taken or even relished by wild birds. This argument is strength- 
 ened by the record of the button-quail {Tumix taigoor) in Finn's 
 experiments. This essentially groimd-loving bird, which is in no 
 way equipped for capturing butterflies under natural conditions, 
 and consequently cannot have an appetite for them, in captivitj' 
 look all but four out of a total of fifty-three that it tried. 
 
 Mason and Lefroy, in the most comprehensive and valuable 
 statement yet published regarding the food of birds in India, say:^ 
 "Butterflies do not form any appreciable proportion of the food of 
 
 ™Finn records the refusal of Aeroea by the red-whiskered bulbul (p. 640), 
 while Poulton {Proc. Enl. Soc. Lond., 1908, p. xxxi) pubUshes a letter from H. L. 
 Andrewes, which states that this bird was observed to feed to its young Acrcea 
 vioUe, supposedly one of the most distasteful of the group. 
 
 " An interesting case of the diversity in results of experiments, and a proof, 
 therefore, of their misleading character, probably refers to this bird, the common 
 bulbul of India. A. G. Butler {Nature, 3, No. 61, December 29, 1870, p. 165) 
 notes that a Mr. Newton, of Bombay, said it was only by repeated persecution 
 that a caged bulbul was induced to touch a Danais. The record of this bird 
 -with Danais in Finn's experiments is A 8 R 4. 
 
 ''Mem. Dept. Agr. India, Ent. Ser., Vol. Ill, January, 1912, p. 338. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 3 2 ) 
 
 any one species of bird, though a good many birds take these insects 
 at times. A long series of experiments with regard to birds taking 
 protectively colored or distasteful insects and especially butterflies 
 
 was made by Mr. Finn They have little importance to 
 
 economic ornithology, since most of the experiments were conducted 
 with caged birds, these, therefore, being under unnatural conditions." 
 
 2. "That many, probably most species, dislike, if not intensely, 
 at any rate in comparison with other butterflies, the "warningly 
 colored" Danainse, Acroea violas, Delias eucharis, and Papilio aristo- 
 lochice; of these the last being the most distasteful and the Danainae 
 the least so" (p. 667). 
 
 By consulting the tabulation of acceptances and rejections given 
 above, it will be seen that only two out of fourteen species of birds 
 considerably experimented with failed to take as many or more 
 insects of the "nauseous" group than they refused, and that seven 
 of these fourteen species refused as large or a larger proportion of 
 the "palatable" butterflies as of the "nauseous" ones. Conse- 
 quently the assertion in Conclusion 2, at least as regards the Dan- 
 ainse, is not borne out even under experimental conditions ; it should 
 read, about half of the species of birds considerably experimented 
 with showed in captivity a greater or less degree of preference for 
 butterflies of the "palatable" group. The figures show that about 
 30 per cent, of all "nauseous" butterflies tested were refused, as were 
 about 13 per cent, of the " palatable " ones. About 23 per cent, of the 
 Danais (Jiverage of three species) and of the Euplceas were rejected, 
 proportions nearly as small or smaller than in the case of at least three 
 species of the so-called palatable group, namely, Papilio demoleus, 25 
 per cent.; Atella phalanta, 22 per cent.; and Elymnias undularis, 24 
 per cent. 
 
 The approximate numbers of refusals and acceptances and the 
 
 percentage of refusals for the important species of both groups are 
 
 given below: 
 
 "Nauseous" group. 
 
 Danais chrysippus 
 
 " genutia 
 
 " limniace 
 
 Delias eucharis 
 
 Euplcea sp 
 
 Euproctis sp 
 
 Mylabris sp... . 
 
 Papilio aristolochicB 
 
 Tenas sp 
 
 Acrcea violce , 
 
 A. 
 
 R. 
 
 %R. 
 
 .36+ 
 
 38 
 
 21.8 
 
 59+ 
 
 23 
 
 28 
 
 38+ 
 
 9 
 
 19.1 
 
 13 
 
 19 
 
 59.3 
 
 53 
 
 16 
 
 22.8 
 
 11 
 
 2 
 
 15.3 
 
 1 + 
 
 1 
 
 50 
 
 17 
 
 33 
 
 66 
 
 11 
 
 4 
 
 26.6 
 
 3 
 
 5 
 
 62.5 
 
6 
 
 22.7 
 
 8 
 
 24.2 
 
 1 
 
 7.1 
 
 1 
 
 7.6 
 
 24 
 
 25.8 
 
 12 
 
 17.& 
 
 330 PROCEEDINGS OF THE ACADEMY OF [JuUe, 
 
 "Palatable" group. 
 
 Atella phalanta 16 
 
 Elymnias undularis 25 
 
 Hypolimnas misippus 13 
 
 Nepheronia hippia 12 
 
 Papilio demoleus 69 
 
 " polites 55 
 
 3. That the mimics of these are at any rate relatively palatable 
 and that the mimicry is commonly effectual under natural condi- 
 tions" (p. 667). 
 
 According to the figures obtained by me, the mimics, Hypolimnas 
 misippus and Nepheronia hippia were each rejected once in thirteen and 
 twelve trials, respectively, an average of about 7 per cent., or much 
 under the average for the "palatable" group as a whole, while about 
 18 per cent, of the Papilio polites and 24 per cent, of the Elymnias 
 undularis were refused, fully as large a proportion as in the case of 
 several members of the "nauseous" group. 
 
 There is no more evidence for the latter half of this conclusion 
 than that any other features of the experiments are analogous to 
 natural conditions. 
 
 As noted above, the experiments and conclusions of Firm are often 
 quoted in support of the selectionist theories, and Finn himself in 
 summing up this earlier work says: "On the whole, the theory of 
 Wallace and Bates is supported by the facts detailed in this and 
 former papers, so far as they deal with birds (and the one- mammal 
 used)" (pp. 667, 668). 
 
 It is of great interest, therefore, to note that the builder of these 
 oft-sought bulwarks of the selectionists later came to the opinion 
 that neither they, nor any other of the defenses brought forward, 
 would save the day for the selection theories. In collaboration with 
 Douglass Dewar, in a book entitled The Making of Species (1909), 
 he says: "Many naturalists, especially Dr. Wallace and Prof. 
 Poulton, have pushed the various theories of animal coloration to 
 
 absurd lengths (p. 171) We have examined these mighty 
 
 images of gold, and silver, and brass, and iron, and found that there 
 is much clay in the feet" (p. 172) .... What we "know of the 
 struggle for existence offers but poor support to the Neo-Darwinian 
 explanation of the cases of the so-called mimicry in nature" 
 (p. 240). 
 
 As a result of his experience with captive birds, Finn recommends 
 that future experimenters use birds in a state of freedom, and at 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 331 
 
 least one experimenter, Lieut.-Col. Neville Manders, has done so. 
 Manders himself says: "I am extremely doubtful as to any real 
 value accruing from experiments on caged birds, whether nestlings or 
 adult. No one, I imagine, believes that all butterflies taste alike; 
 no doubt some are more tasty than others, and caged birds fed upon 
 butterflies, even with other insect food, would no doubt learn in 
 time to distinguish the different kinds; but this procedure to my mind 
 begs the question, as it assumes that butterflies are an ordinary 
 article of food in the wild state, a proposition .... which the 
 evidence .... does not altogether support."'^ It is noteworthy 
 that the free birds Manders did induce to take disabled butterflies 
 were not seen by him to attack these insects under normal con- 
 ditions. He frequently comments (pp. 736-739, 741) on this fact. 
 Although the birds ate the helpless butterflies, they took no notice 
 of the freely flying ones that abounded in the vicinity. 
 
 The wild birds experimented upon in Ceylon by Manders, with 
 their records, are as follows: (Disregarded — D. — means simply 
 not taken and not tried. Behavior toward dead butterflies not 
 noted). 
 
 Nauseous Palatable 
 
 group. group. 
 
 A. R. D. A. R. D. 
 
 Robin flycatcher, Siphia hyperythra 2 2 6 1 
 
 Dusky-blue " Stoparola sordida 7 2 4 
 
 Brown shrike, Lanius cristatus 4 7 4 5 
 
 Magpie robin, Copsychus saularis... 10 3 44- 21 -|- 6 
 
 Mynah, Acridotheres tristis.. . 11 1 5 1 
 
 34 3 14-1- 38+ 17 
 
 Thus there were no refusals (upon trial) of any living butterflies 
 except by the magpie robin. This bird has three rejections, two of 
 Euplcea core, one of which it ate immediately afterwards. The 
 bird's record with this butterfly was A 9 R 2. Manders says the 
 other butterfly (Terias hecabe) refused by this species was too dry. 
 The percentage of insects disregarded is practically the same for the 
 " nauseous " and the " palatable " groups. Manders' conclusion from 
 this and other evidence is that "the terms palatable and unpalatable 
 are not justified at present" (I.e., p. 742). 
 
 Experiments in America. 
 Unfortunately, the natural food habits of many of the Indian, 
 
 " Proc. Zool. Soc. Lond., September, 1911, p. 745. 
 
332 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 African, and British birds experimented with are not well known, 
 for the selectionists have exaimined very few stomachs of wild birds. 
 This method is more arduous and does not pile up results so hand- 
 somely as do experiments. But it is, nevertheless, in connection 
 with the strictly correlated examination of contents of other parts 
 of the alimentary canal, and of pellets, and fseces (together with 
 reUable records of individuals seen or collected with food in talon or 
 beak), the only trustworthy method of learning what birds actually 
 eat under natural conditions. And this information only is accep- 
 table proof of the -tastes and food preferences of birds or, for that 
 matter, of any other animals. 
 
 It is fortimate, therefore, that one series of experiments has been 
 made the results of which can be closely checked with a satisfactory 
 amount of exact information upon the food habits of the same species 
 under natural conditions. 
 
 Experiments by Judd and Beat. 
 
 The experiments referred to have never been published upon as a 
 whole, though some of the results may be found in the following 
 publications: 
 
 Beal, F. E. L. The Bluejay and its Food. Yearbook U. S. Dept. 
 
 Agr., 1896 (1897), pp. 205, 206. 
 Birds of California in relation to the Fruit Industry. Part I, 
 
 Bui. SO, Biol. Survey, 19a7, p. 35. 
 Judd, Sylvester D. Four common Birds of the Farm and Garden. 
 
 Yearbook U. S. Dept. Agr., 1895 (1896), pp. 410, 414. 
 The Efficiency of some Protective Adaptations in securing 
 
 Insects from Birds. Am. Nat., 33, No. 390, June, 1899, pp. 
 
 461^84. 
 The relation of Sparrows to Agriculture. Bui. 15, Biological 
 
 Survey, 1901, pp. 45-48. 
 The Bobwhite and other Quails of the United States in their 
 
 economic relations. Bui. 21, Biological Survey, 1905, pp. 
 
 28, 29, 36, 38, 40, 41, 44-45. 
 
 Doctor Judd was at one time very enthusiastic with regard to 
 experiments in feeding birds, and these experiments were initiated 
 and largely carried on by him. They were watched, however, and 
 in part performed by Prof. F. E. L. Beal, the veteran economic 
 ornithologist, who has examined the contents of more bird stomachs 
 than any other person in the world. Prof. Beal was mainly respon- 
 sible for the discontinuance of these experiments, and I am betraying 
 no secret in asserting that experimental ornithology was abandoned 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. . 333 
 
 by the United States Biological Survey because of a direct realization 
 from these trials of the futility of experiments as indications of the 
 food preferences and, therefore, of the economic status of species 
 under natural conditions. 
 
 It is not the writer's purpose to give a detailed account of these 
 experiments, but merely lists of the items accepted and rejected, with 
 comments thereon. It will be helpful to consider separately those 
 items which were both refused and devoured. Several discrepancies 
 exist between the statistics here presented and the published accounts 
 previously referred to, but the writer has made the following tabula- 
 tions directly from notebooks containing daily entries regarding 
 the experiments. He believes these should be accepted as correct, 
 rather than statements in the printed pages that have run the gantlet 
 of editors and proof-readers, whose efforts often have just the oppo- 
 site result, so far as accuracy is concerned, from that which the 
 exercise of their true functions is intended to insure. 
 
 To interpret the bearing of this and the following experiments on 
 the theory of protective adaptations, it should be recalled that the 
 common types of what is called warning coloration are the combina- 
 tions of black with red, yellow, and white. Metallic colors also are 
 usually classed as warning. Besides the insects, etc., possessing 
 these colors, other groups, for various reasons, are said to be specially 
 defended. Among these are ground beetles (Carabidae), many of 
 which have acid and nauseous secretions; the true bugs (Hemiptera), 
 nearly all pungently flavored and malodorous; ants, and the stinging 
 wasps and bees (Hymenoptera) ; the spiders and centipeds with 
 poison fangs; and the millipeds with acid juices. All of these crea- 
 tures are supposed to be especially protected from the attacks of 
 predaceous animals or, in other words, to be distasteful. 
 
 To bring out clearly the attitude of Judd's captive birds toward 
 these categories of "protected" animals, the writer has tabulated 
 the results (as regards the animal food only) of each series (except 
 the shorter ones) of experiments under the following headings: 
 "warningly colored" species, others "specially defended," and 
 "non-protected" species. Of course, the term "non-protected" 
 is not in accordance with the theories of protective adaptations, as 
 the more obscurely colored and innoxious forms thus described are 
 also said to be protected, but chiefly in a more passive way than the 
 other two groups, namely, by concealing coloration. "Non-pro- 
 tected" is therefore used to bring into greater contrast the theoretical 
 attributes of these comparatively poorly "protected" species. 
 
334 
 
 PROCEEDINGS OF THE ACADEMY OF 
 
 [June, 
 
 BoBWHiTE {Colinus virginianus) -P — 
 Accepted: 
 
 Ch^topoda. 
 
 Earthworms. 
 Coleoptera. 
 Carabid^. 
 
 Scarites suhterraneus (black). 
 
 Harpalus erythropus (black, reddish legs). 
 
 CoCCINELLIDiB. 
 
 AdaJia bipunctata (red, black, and yellow). 
 Epilachna borealis (yellow and black). 
 
 ScaRAB^IDjE. 
 
 Ligyrus gibbosus (red-brown). 
 
 ChRYSOMELIDjE. 
 
 Diabrotica 12-punctata (yellow and black), 2+. 
 Diabrotica vittata (yellow and black), 2+. 
 Leptinotarsa decemlineata (yellow and black), three birds 
 ate fifty in five minutes. 
 Lepidoptera. 
 
 PlERIDiE. 
 
 Pieris rapce larvae (green, black, and yellow), 2. 
 Sphingid^. 
 
 Phlegethontius sp. larvse (green and white), 2. 
 
 NOCTUID^. 
 
 Agrotis sp. larva. 
 Rejected: 
 
 Coleoptera. 
 coccinellid^. 
 
 Chilocorus bivulnerus (black. and red). 
 
 !MeLOIDjE. 
 
 Melee angusticollis (dark blue or violet, vesicant body 
 fluids). 
 
 HOMOPTERA. 
 APHID.E. 
 
 Hymenoptera. 
 
 TenthredinidjB, larvae. 
 
 Summary: Colinus virginianus. — 
 
 
 Accepted. 
 
 Rejected. 
 
 * 
 
 Species.' ^Plt 
 '^ 1 mens. 
 
 Species. 
 
 Speci- 
 mens. 
 
 "Warningly colored" species 7 59^ 
 
 Others "specially defended" i i 
 
 "Non-protected" species j 4 j 5 
 
 2 
 2 
 
 
 2 
 
 2 + 
 
 
 ^ Partial account of these experiments in Bui. 21, Biol. Survey, 1905, pp. 28-29, 
 36, 38, 40, 41, 44-45. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 335 
 
 Thus these bobwhites ate, among other things, three species of 
 strongly flavored yellow and black Chrysomelidse, or leaf beetles, 
 and two species of equally if not more pungent Coccinellidse or 
 ladybirds, whose colors of yellow and black and red, black, and 
 yellow are typically "warning." On the other hand, the birds 
 refused one red and black ladybird. It is evident considerations as 
 to color of prey have little weight with the quail. It is worthy of 
 note also that although these birds refused plant lice, birds experi- 
 mented upon by Mrs. Margaret M. Nice ate large numbers of these 
 insects. 
 
 Mrs. Nice's experiments upon bobwhites*' which have previously 
 been reviewed'^ by the writer clearly bring out the fact that birds 
 will eat in captivity insects which they probably never eat or in 
 some cases never even see in their normal existence. Examples are: 
 house-flies (Musca domestica) and mosquitoes; 1350 and 568 of 
 these insects, respectively, were taken at single meals, but undoubtedly 
 they are seldom if ever eaten by wild bobwhites. Silver fish (Lepisma 
 saccharina), clothes moths {Tinea pellionella), and mealworms 
 (Tenebrio) also were eaten by the captive quail, but wild birds 
 probably never have a chance to get these close associates of man. 
 
 The writer does not list the results of Judd's trials of quail with 
 various vegetable foods, but only one item was refused, namely, 
 strawberries. These are eaten by wild bobwhites and Judd com- 
 ments'^ on the fact as follows: "M. B. Waite reports that near 
 Odenton, Md., it sometimes picks ripening strawberries. Yet 
 birds that were kept in captivity several months refused straw- 
 berries when they were hungry." 
 
 Begad WINGED Hawk (Buteo platypterus) . — 
 
 Accepted: 
 
 Lepidoptera. 
 
 Basilona imperialis imago (yellow and purplish-brown). 
 Bartrachia. 
 
 Bujo sp. 
 
 AVES. 
 
 Micropodid^. 
 
 Chwtura pelagica (fuscous). 
 
 " "Food of the Bobwhite." By Margaret Morse Nice, Journ. 0/ Economic 
 Entomology, Vol. 3, No. 3, June, 1910, pp. 295-313. 
 
 " Journ. Economic Ent., Vol. 3, No. 5, October, 1910, pp. 437-438. 
 » Bui. 21, Biol. Survey, 1905, p. 36. 
 
336 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 FeINGILLIDjE. 
 
 Passer domesticus (nestlings), 3. 
 
 There is no record of a refusal by this bird. The toad is supposed 
 to be protected by acrid secretions of glands in the skin. 
 
 RuBT-THROATED HUMMINGBIRD {ArchUochus coluhrls) : — 
 
 Rejected: 
 
 Small Aphidse. 
 " Jassidae. 
 " Culicidse. 
 " Other Diptera. 
 " Halticinae. 
 " Araneida. 
 
 The leaf hoppers (Jassidse), small flies (Diptera), flea-beetles 
 (Halticinae), and spiders (Araneida), at least, are common articles in 
 the natural diet of this species. 
 
 Bluejay {Cyanocitta cristata) :^ — 
 
 Accepted: 
 
 Ciletopoda. 
 
 Earthworms, 7. 
 Isopoda. 
 
 Oniseus asellus, 6. 
 Chilopoda. 
 
 Lithobius sp., 2. 
 
 Julus sp., 2. 
 Ephemerida. 
 
 Adults, many. 
 Orthopteea. 
 Acridiid^. 
 
 Dissosteira Carolina, 1. 
 
 COLEOPTERA. 
 CaEABIDjE. 
 
 Agonoderus palKpes (pale yellow and black), 1. 
 Anisodactylus discoideus (black and brownish-yellow), 2. 
 
 " rusticus (brownish-black), 2. 
 
 Calosoma scrutator (metallic green, red, and blue), 1. 
 Chlcmius sp., 1. 
 
 Gakrita janus (black and reddish-brown), 2. 
 Harpalus caliginosus (black), 1. 
 
 »« Partial account in Yearbook U. S. Dept. Agr., 1896 (1897), pp. 205, 206. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 337 
 
 ELATERID.E. 
 
 Alaus oculatus (black and silvery, with eye-spots). 
 Elaterid, adult, 1. 
 Elaterid, larva, 1. 
 Scarab^idjE. 
 
 Allorhina nitida (green and yellow), 1. 
 
 LUCANID^. 
 
 Passalus cornutus (black), 2. 
 
 CERAMBYCIDiB. 
 
 Typocerus sinuatus (black with yellow bands), 5. 
 Chrysomelid^. 
 
 Diabrotica 12-punctata (yellow with black spots), 1. 
 Tenebrionid^. 
 
 Nydobates pennsylvanicus (black), 1. 
 Tenebrio obscurus (dark reddish-brown), 1. 
 Tenebrionid undet., 1. 
 Heteroptera. 
 Pentatomid^e. 
 
 Brochymena sp., 1. 
 Lepidoptera. 
 
 Philosamia cynthia (yellow and purplish -brown), 3 (alive- 
 
 and dead) . 
 Telea polyphemus ad. (mainly reddish-brown, white and. 
 
 black, eye spots on each pair of wings), 1. 
 Orgyia leucostigma, pupse, 2. 
 Hyphantria cunea, larvse (very hairy), many. 
 Hairy caterpillar, undet., 1. 
 Cutworm, 1. 
 Hymenoptera. 
 Apina, 2. 
 
 Agapostemon sp. (metallic green), 1. 
 Araneida. 
 
 Spider, 1. 
 
 AVES. 
 
 Phasianid^. 
 
 Egg shells. 
 Fringillid^. 
 
 English sparrow eggs, 2. 
 
 Rejected: 
 
 Orthoptera. 
 Blattid^. 
 
 Stylopyga orientalis (black). 
 Heteroptera. 
 Belostomatid^. 
 
 Benacus griseus (light brown). 
 
338 proceedings of the academy of [junc, 
 
 coleoptera. 
 Lampyrid^. 
 
 Chauliognathus pennsylvanicus (yellow and black), 3. 
 Chrysomelid^. 
 
 Chrysochus auratus (metallic green and coppery). 
 Lepidoptera. 
 Papilionid^. 
 
 Papilio troilus adult (dark red-brown, white, and bluish- 
 green). 
 
 PULMONATA, 
 LiMACIDJE. 
 
 Ldmax sp. 
 
 AVES. 
 
 PhASIANIDjE. 
 
 Hen's egg (whole). 
 Feingillid^. 
 
 Passer domesiicus (alive), in cage three days. 
 Mammalia. 
 Murid^. 
 
 Mus musculus (alive). 
 Spermatophyta. 
 Morace^. 
 
 Morus sp. (berry), 2. 
 
 AqUIFOLIACRffi. 
 
 Ilex opaca, berry (red), 2. 
 Disregarded: 
 Coleoptera. 
 Coccinellid^. 
 
 Adalia bipundata (red, black, and yellow), 2. 
 Hippodamia sp., 1. 
 Chrysomelid-e. 
 
 Galerucella luteola (yellow and black), 3. 
 Lepidoptera. 
 Papilionid^. 
 
 Papilio turnus ad. Killed, dropped when frightened by 
 • observer, not picked up (mainly yellow and black), 1. 
 Hymenoptera. 
 Apina. 
 
 Apis mellifera, worker (brown), 1. 
 Spermatophyta. 
 Fagace^b. 
 
 Fagus grandifolia, nuts. 
 
 MyRTACEjE. 
 
 Citrus sp., whole fruit. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 339 
 
 Accepted and Rejected: 
 Orthoptera. 
 Gryllid^. 
 
 Gryllus sp., A 1 R 1. 
 coleoptera. 
 Carabid^e. 
 
 Scarites subterraneus (black), A 2 R 1. 
 Hydrophilid^. 
 
 Hydrophilus triangularis (shining greenish-black), A 2 R 2. 
 
 SCARAB^ID^. 
 
 Ldgyrus gibbosus (reddish-brown), A 17 R 2. 
 Hymenoptera. 
 
 Bombus sp., A 1 R 1. 
 
 Xylocopa virginica, worker (black with yellow hairs), 
 A 1 R 1, male A 1. 
 Mammalia. 
 
 MuRIDjE. 
 
 Mus musculus (dead), A 3 R 1. 
 Accepted and Disregarded: 
 Coleoptera. 
 
 ScARABiEID^. 
 
 Lachnosterna sp. (reddish-brown), A 3 D 2. 
 Hymenoptera. 
 Myrmicid^b. 
 
 Tetramorium ccespitum (a minute reddish ant), A 1, D many. 
 
 Accepted, Rejected, and Disregarded: 
 Orthoptera. 
 Blattid^. 
 
 Blattella germanica (yellow-brown and dark brown), 
 A 16-1- Rl Dl. 
 Coleoptera. 
 Scarab^id^. 
 
 Dysdnetus trachypygus (black), A 5 R 2 D 2. 
 Lepidoptera. 
 
 Philosamia cynthia, cocoons. One pecked, could not be 
 opened, was left; two others disregarded were after- 
 wards eaten when cut open. 
 Spermatophyta. 
 Fagace^. 
 
 Quercus sp. (acorns), A 8-1- D several, R. 
 
 Disregarded and Refused: 
 Coleoptera. 
 
 Leptinotarsa decemlineata (yellow and black), D 1 R 2. 
 Lepidoptera. 
 
 Colias philodice ad. (yellow and black), D 2 R 1. 
 
340 PROCEEDINGS OF THE ACADEMY OF 
 
 Summary: Cyanocitta cristata. — 
 
 [June, 
 
 Accepted. Rejected. Disregarded. 
 
 Spe- Speci- Spe- Speci- j Spe- j Speci- 
 cies. mens. cies. mens. ■ cies. ' mens. 
 
 "Wamingly colored" species. 
 Others "specially defended". 
 "Non-protected" species 
 
 12 
 
 8 
 
 15 
 
 21+ 3 
 12 1 
 
 29 5 
 
 4 7 
 
 1 1 
 
 
 
 "Wamingly colored" species 
 Others "specially defended". 
 "Non-protected" species 
 
 Accepted and 
 rejected. 
 
 Accepted and 
 disregarded. 
 
 cies" Specimens. V^^ Specimens. 
 
 A3R2 AODO 
 2 1 1 1 H 
 
 23 6 13 2 
 
 Accepted, rejected I Disregarded and 
 and disregarded, l rejected. 
 
 cfes' I Specimens. I ^P^' Specimens. 
 
 "Warningly colored" species.. 
 Others "specially defended"... 
 ' ' Non-protected " species 
 
 
 
 JAORODO 
 
 2 
 
 D3 
 
 R3 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 24+ 4 5 
 
 
 
 
 
 
 
 Images of Philosamia cynthia, at least, among the things ac- 
 cepted are seldom or never encountered by wild bluejays. This 
 species was imported with its food plant Ailanthus glandulosus, and 
 is established in very few places. The cocoons of this species could 
 not be opened by the jay, but when opened for him the pupae were 
 eaten. Of the items refused, hen's eggs are all too often attacked 
 by wild birds; living birds and mice are frequently killed and eaten, 
 and mulberries also are eaten under natural conditions. 
 
 Among things both accepted and rejected, crickets (Gryllus) and 
 May-beetles (Lachnosterna) are commonly devoured by free birds. 
 The carabid Scarites also has been found in the stomach of this 
 species. Beechnuts were disregarded by the captive jay, and acorns 
 were both disregarded and refused, though some were afterwards 
 eaten. Both of these nuts are frequently eaten by wild jays. These 
 instances in themselves are sufficient to show that acceptances and 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 341 
 
 rejections by captive birds are no guide to the natural tastes of the 
 species. 
 
 English Sparrow {Passer domesticus) : — 
 Accepted: 
 
 COLEOPTERA. 
 
 ScARABjEIDjE. 
 
 Ligyrus gibbosus (reddish-brown), 1. 
 Spermatophyta. 
 Gramin^. 
 
 Panicum sanguinale. 
 Chcetochloa italica. 
 Chcetochloa viridis. 
 
 .ChENOPODIACEjE. 
 
 Chenopodium album. 
 
 AmBROSIACEjE. 
 
 Ambrosia artemisiwfolia. 
 Rejected: 
 
 Heteroptera. 
 Pentatomid^e. 
 
 Brochymena arborea (dark brown), 1. 
 Spermatophyta. 
 
 CiCHORIACEiE. 
 
 Taraxacum taraxacum (heads with akenes). 
 
 Accepted and Rejected: 
 Spermatophyta. 
 Amaranthace^. 
 
 Amaranthus sp. Refused at 9 A.M. when hungry, but 
 eaten at 11 same morning. 
 
 The fruiting heads of dandelion which were refused are a favorite 
 natural food; and Amaranthus seeds, which were refused but eaten 
 two hours later, are commonly eaten by \\Tild birds of this species. 
 
 Snowbird {Junco hyemalis) :^' — 
 
 Accepted: 
 
 Orthoptera. 
 
 Encoptolophus sordidus (brown). 
 Rejected: 
 
 Coleoptera. 
 coccinellid^. 
 
 Adalia bipunctata (red, yellow, and black). 
 
 SCARAB^ID^. 
 
 Dysdnetus trachypygus (black). 
 
 " Partial account of experiments with this and the following two species in 
 But. 16, Biol. Survey, 1901, pp. 45-48. 
 
342 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 Chrysomelid^. 
 
 Lema trilineata (yellow and black). 
 Meloid^. 
 
 Epicauta sp. 
 Spermatophyta. 
 Chenopodiace^ . 
 
 Chenopodium sp. (seeds). 
 
 The latter seeds are a common natural food. 
 White-throated Sparrow {Zonotrichia albicolKs) : — 
 
 Accepted: 
 
 Heteroptera. 
 Pentatomid^. 
 
 Murgantia histrionica (orange and black). 
 
 Other acceptances and rejections same as with Junco, and same 
 remark applies. 
 
 Song Sparrow (Melospiza melodia) : — 
 
 Accepted: 
 
 Neuroptera. 
 
 ChRYSOPIDvE. 
 
 Chrysopa sp., 1. 
 Orthoptera. 
 
 Encoptolophus sordidus (brown), 1. 
 Coleoptera. 
 Carabid^. 
 
 Amara sp., 2. 
 
 Anisodactylus terminatus {dark brown to greenish-black), 1. 
 
 Harpalus pennsylvanicus (black), 1. 
 
 Nebria pallipes (black, legs yellow), 1. 
 
 Platynus sp., 2. 
 
 Pterostichus sayi (green), 3. 
 Trogositid^. 
 
 Trogosita virescens (metallic green or blue), 1. 
 Curculionid^. 
 
 Sitones sp., 2. 
 
 HOMOPTERA. 
 
 JassidjE (nymphs and adults), 3. 
 Heteroptera. 
 
 Lyg^id^b, 1. 
 
 Reduviid^, 1. 
 Lepidoptera. 
 
 PaPILIONIDjB. 
 
 Papilio turnus, ad. (mainly yellow and black), 1. 
 Moths, 2. 
 Araneida. 
 
 Spider, 1. 
 
1912.] natural sciences of philadelphia. 343 
 
 Spermatophyta. 
 
 CARYOPHYLLACEiE. 
 
 Alsine media (seeds). 
 
 Rejected: 
 
 coleoptera. 
 Carabid^. 
 
 Various Harpini were refused, but later Harpalus penn- 
 sylvanicus was eaten. 
 
 COCCINELLID^. 
 
 Adalia bipundata (red, black, and yellow), 1. 
 Hippodamia sp., 2. 
 
 ScARABiBID^. 
 
 Allorhina nitida (green and yellow).. 
 Lachnosterna sp. 
 
 ChRYSOMELIDjE. 
 
 Diabrotica 12-punctata (yellow and black), several. 
 
 Lema trilineata (yellow and black). 
 Meloid^. 
 
 Epicauia sp. 
 Lepidoptera. 
 Arctiid^. 
 
 Leucarctia acrcea, ad. (white, yellow, and black). 
 Hymenoptera. 
 formicid^. 
 
 Black ant, probably Camponotus. 
 Spermatophyta. 
 Chenopodiace^. 
 
 Chenopodium sp. 
 Polygonace.*;. 
 
 Polygonum sp. 
 
 Accepted and Rejected: 
 
 CoLEOPTERA. 
 
 Carabid.*;. 
 
 Agonoderus pallipes (pale yellow and black), ate 2, refused 
 
 others, but next day ate 15 in three minutes. 
 Chlcenius sp., A 1 R 1. 
 
 SCARAB^ID^. 
 
 Dyscinetus trachypygus (black), A 1 R 1. 
 Trox sp., A 1 R 2. 
 Spermatophyta. 
 Amaranthace^e. 
 
 Amaranthus sp. Refused at first, finally starved into 
 eating it. 
 
 ClCHORIACE^. 
 
 Taraxacum taraxacum. The opened fruiting head was 
 accepted at the only trial, the closed involucres were 
 at first refused; several days afterwards 3 were eaten. 
 
•344 PROCEEDINGS OF THE ACADEMY OF 
 
 Summary: Melospiza m^lodia. — 
 
 [June, 
 
 
 Accepted. 
 
 Rejected. 
 
 Accepted and 
 rejected. 
 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 Spe- 
 cies. 
 
 Specimens. 
 
 "Warningly colored" species.... 
 
 Others "specially defended" 
 
 "Non-protected" species 
 
 4 
 9 
 3 
 
 6 
 
 13 
 
 5 
 
 5 
 
 3 + 
 2 
 
 6 
 
 3 + 
 2 
 
 2 
 
 2 
 
 A 18 R2-I- 
 
 
 2 3 
 
 
 
 
 In rejecting the seeds of Chenopodium and Polygonum this bird 
 jef used two favorite items of the food of wild members of its species ; the 
 bird was only starved into eating Amnranthus seeds, another favorite 
 natural food. Of the rejected insects, Camponotus, Diabrotica, Hippo- 
 damia, and Lachnosterna have been found in collected stomachs. The 
 Acceptances include at least one insect, Papilio turnus, which the 
 bird probably never gets under natural conditions. The experi- 
 menter noted that this butterfly would have easily escaped the bird 
 liad it not been confined. 
 
 Little Butcherbird (Lanius ludovicianus) :** — 
 
 Accepted: 
 
 Chilopoda. 
 
 Ldthobius sp., 1. 
 Orthoptera. 
 acridiid.e. 
 
 Hippiscus sp., 2. 
 Coleoptera. 
 scarab.eid^. 
 
 Copris Carolina (black), 1. 
 Lachnosterna sp., 1. 
 
 Ldgyrus gibbosus (dead) (reddish-brown), 1. 
 Osmoderma sp., 1. 
 
 Trichius piger (greenish-black, reddish-brown, white; both 
 white and yellow hairs), 1. 
 
 CERAMBYCIDiE. 
 
 Monohammus sp., 1. 
 
 MELOID.E. 
 
 Meloe americana (bluish-black, vesicating juices), 1. 
 Hetehoptera. 
 
 PeNTATOMIDjE. 
 
 Euschistus sp., 1. 
 Nezara hilaris (green), 1. 
 
 »* Partial account in Bui. SO, Biol. Survey, 1907, p. 35. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 345 
 
 BeLOSTOMATIDjE. 
 
 Belostoma americanum (light brown), 1. 
 Lepidoptera. 
 
 Ceratomia catalpce larva (black and yellow), 2. 
 Estigmene acrcea, ad. (white, yellow, and black), 2. 
 Euvanessa antiopa, ad. (reddish-brown, light blue, and 
 yellow), 2. 
 
 DiPTERA. 
 
 Calliphora sp., 2. 
 Pisces. 
 
 Goldfish, 1. 
 
 Micropterus salmoides, 1. 
 Urodela. 
 
 Plethodon cinereus erythronotus (brown and red). 
 
 Plethodon glutinosus (black and white), 1. 
 Reptilia. 
 
 Storeria dekayi (grayish-brown and black), 1. 
 
 Heterodon platyrhinus (yellow or reddish and brown or 
 black), 1. 
 
 Sceloporus undulatus (gray or brown and black), 1. 
 
 AVES. 
 
 FRINGILLIDiB. 
 
 Passer domesticus, 3. 
 
 ViREONIDiE. 
 
 Vireo olivaceus, 4. 
 
 TrOGLODYTIDjB. 
 
 Telmatodytes palustris (dead), 1. 
 Mammalia. 
 
 Mus musculus, 5. 
 
 Rejected: 
 
 coleoptera. 
 . Erotylid^. 
 
 Megalodacne heros (black and yellow), 1. 
 
 LaMPYRIDjE. 
 
 Chauliognathus pennsylvanicus (yellow and black), 1. 
 Chrysomelid^. 
 
 Diabrotica 12-piuictata (yellow and black), 4. 
 Meloid^. 
 
 Epicauta vittata (yellow and black), 1. 
 Heteroptera. 
 
 PeNTATOMIDjE. 
 
 Murgantia histrionica (yellow and black), 1. 
 
 Lepidoptera. 
 
 Euvanessa antiopa larva (black, spiny), 1, pupa 1. 
 Hyphantria cunea larva (yellow, brown, and black, very 
 
 hairy), 2. 
 Malacosoma americana larva (black, white, and blue, 
 
 hairy), 1. 
 
346 
 
 PROCEEDINGS OF THE ACADEMY OF 
 
 [June, 
 
 Orgyia leucostigma larva (red, black, white, and yellow, 
 hairy tufted), 1. 
 Spermatophyta. 
 R0SACE.1B. 
 
 Fragaria sp. (fruit). 
 
 Accepted and Refused: 
 
 COLEOPTERA. 
 CARABID.ffi. 
 
 Calosoma scrutator (metallic blue, red, and green). The 
 butcherbird seemed to be staggered by the effluvium 
 of one of the first Calosomas given, but devoured it. 
 Later one was offered it rear end first through the wall 
 of the cage; the bird ate part of the viscera, then 
 refused to touch it again. However, he ate two the 
 next day. In all seven were devoured. 
 
 SlLPHID^. 
 
 Silpha incequalis (black), A 1 R 1. 
 
 SCARAB^ID^. 
 
 Trox sp., A 1 R 1. Both the Silpha and the Trox were 
 refused in presence of experimenter, but their remains 
 were found in a pellet thrown up by the bird about two 
 hours later. 
 Hymenoptera. 
 Apina. 
 
 Apis mellifera, worker (brown), A 1 R 1; drone, A 1. 
 Mammalia. 
 MuRiD.^;. 
 
 Mus norvegicus, A 2 R 2. 
 
 It is difficult to believe that a bird acting on principle would 
 refuse Epicauta vittata and eat Meloe americana. Diabrotica is eaten 
 by wild individuals of the species, and Silpha'which was both accepted 
 and rejected by this bird, is a common capture. Calosomas also 
 are frequently eaten by wild butcherbirds. 
 
 Summary: Lanius ludovidanus. — 
 
 
 
 
 
 
 
 Accepted. 
 
 Rejected. 
 
 Accepted and 
 rejected. 
 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 "Waxningly colored species" 
 
 12 
 
 4 
 
 11 
 
 16 
 
 4 
 
 21 
 
 8 
 1 
 
 
 12 
 2 
 
 
 1 
 2 
 2 
 
 A7 Rl 
 
 Others "specially defended" 
 
 "Non-protected species" 
 
 3 2- 
 3 3 
 
 Mockingbird (Mimus polyglottos). — 
 
 Twice chose the grasshopper {Encoptolophus sordidus) in prefer- 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 347 
 
 ence to the May-beetle {Lachnosterna) , although the latter is eaten 
 by wild mockingbirds. 
 
 Catbird {Dumetella carolinensis) :" — 
 Accepted: 
 
 CHiETOPODA. 
 
 Earthworm. 
 
 ISOPODA. 
 
 Oniscus asellus, 6. 
 Chilopoda. 
 
 Julus sp. 
 Orthoptera. 
 
 Green Acridiid. 
 coleoptera. 
 Carabid^. 
 
 Undetermined, 11. 
 
 Agonoderus pallipes (pale yellow and black), 1. 
 
 Anomoglossus pusillus (bluish-green, blue, or black). 
 
 Bemhidium chalceum (coppery to greenish or black). 
 
 STAPHYLINIDiB. 
 
 Undet. (with red elytra), 1. 
 
 CoCCINELLIDiE. 
 
 Undet., 1. 
 CucujidjE. 
 
 Cucujus clavipes (larvse), 6. 
 Dbrmestid^. 
 
 Dermestes talpinus (black with variously colored hairs), 4. 
 
 ScARABjBIDiB. 
 
 Lachnosterna sp. 
 Onthophagus hecate (black). 
 
 CHRYSOMELIDiB. 
 
 Diabrotica 12-punctata (yellow and black). 
 Tenebrionid^. 
 
 Tenebrio molitor (reddish-brown or black), 2. 
 
 CURCULIONID^. 
 
 Centrinus scutellum-album (gray). 
 Lepidoptera. 
 
 Erwetnessa antiopa, ad. (reddish-brown, light blue, and 
 
 yellow). 
 Phlegethontius 5-maculatus, ad. (gray, dark brown, and 
 
 yellow). 
 Hyphantria cunea, larva (yellow, brown, and black, very 
 
 hairy), 5. 
 
 DiPTERA. 
 
 Calliphora erythrocephalus, larvae, 56; ad. (metallic blue, 
 
 eyes dull red), 1. 
 Musca domestica (black and gray), 2. 
 
 «> Partial account in Yearbook U. S. Dept. Agr., 1895 (1896), p. 410. 
 
348 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 Hymenoptbra. 
 formicoidea, 7. 
 Lasius alienus, 22. 
 Cremastogaster linearis, pupa, 3. 
 . Formica sp., 7. 
 Camponotus pennsylvanicus, 8. 
 Araneida. 
 
 Lycosa sp. 
 
 PULMONATA. 
 
 Limax sp., 3. 
 Spermatophyta. 
 Horaces. 
 
 Morus sp., many. 
 
 Rejected: 
 
 Coleoptera. 
 buprestid.*;. 
 
 Undet., 3 (could not break the insects). 
 Hymenoptera. 
 Wasp. 
 
 Disregarded: 
 Coleoptera. 
 
 Passalus cornutus (black). 
 Lepidoptera. 
 
 Euvanessa antiopa, larva (black, spiny). 
 
 Accepted and Rejected: 
 Coleoptera. 
 Carabid.*;. 
 
 Chloenius sp., A 3 R 5. 
 
 TENEBRIONIDiE. 
 
 Nyctobates pennsylvanicus (black), A 3 R 3. 
 Hymenoptera. 
 Apina. 
 
 Apis mellifera, workers (brown), A 5 R 2. 
 
 PULMONATA. 
 
 Snail (large), R 1; (small), A 2. 
 Spermatophyta. 
 Rosacea. 
 
 Strawberry (Fragaria sp.), A several, R several. 
 Amygdalace^b. 
 
 Cherry (Cerasus sp.), A 2 R several. 
 
 Accepted and Disregarded: 
 Coleoptera. 
 
 LaMPYRUJjE. 
 
 Chauliognathus pennsylvanicus (yellow and black), D 1 
 Al. 
 
 Of the insects accepted, the larvae of Cucujus davipes, at least, 
 
1912.] 
 
 NATURAL SCIENCES OF PHILADELPHIA. 
 
 349 
 
 which live under the bark of rotting trees, are probably never en- 
 countered by wild catbirds, but all offered the captive birds were 
 eaten. Among the forms accepted and rejected, the nauseous 
 metallic-green ground-beetle {Chloenius) and the honey-bee (Apis 
 melUfera) are eaten by wild catbirds, and both strawberries and 
 cherries are favorite foods, as cultivators to their sorrow well know. 
 A wasp was rejected by the captive birds, but many wasps have been 
 found in stomachs of wild catbirds. The soldier-beetle (Chauliogna- 
 thus pennsylvanicus) , both disregarded and accepted in the experi- 
 ment, is eaten under natural conditions. 
 
 Summary: Dumetella carolinensis. — 
 
 
 Accepted. 
 
 Rejected. 
 
 Disregarded. 
 
 * 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 Spe- 1 Speci- 
 cies. 1 mens. 
 
 "WnTTlingly nolnrpH " spp(<ips 
 
 9 
 
 16 
 62 
 80 
 
 
 1 
 1 
 
 
 1 
 3 
 
 
 
 Others "specially defended" 
 
 10 
 11 
 
 1 1 
 
 "Non-protected" species 
 
 1 1 
 
 Accepted and 
 rejected. 
 
 Accepted and 
 disregarded. 
 
 Spe- 
 cies. 
 
 "Wamingly colored" species. 
 Others "specially defended". 
 "Non-protected" species 
 
 Specimens. 
 
 Spe- 
 
 Specimens. 
 
 Brown Thrasher (Toxostoma rufum) •?" — 
 
 Accepted: 
 
 coleoptera. 
 
 Carabid^. * 
 
 Harpalus caliginosus (black). 
 
 LaMPYRIDjE. 
 
 Chauliognathus pennsylvanicus (yellow and black). 
 
 ChRYSOMELIDjB. 
 
 Leptinotarsa decemlineata (yellow and black), twice swal- 
 lowed and thrown up, then swallowed again and re- 
 tained. 
 
 Diabrotica 12-punctata (yellow and black). 
 
 ™ An account of these experiments, with some additional remarks about the 
 bird's preferences for certain human foods and wild berries, is in Yearbook U. S. 
 Dept. Agr., 1895 (1896), p. 414. 
 
350 proceedings of the academy of [junc, 
 
 Meloid^. 
 Epicauta sp. 
 Hetebopteha. 
 
 CoKEIDjB. 
 
 Anasa tristis (brown). 
 Lepidopteka. 
 
 PiEHIDjB. 
 
 Pieris rapce, larvae (green, yellow, and black), 3. 
 
 AeCTIIDjB. 
 
 Caterpillar, swallowed, but throT^n up. 
 
 Rejected: 
 
 Lepidopteka. 
 
 Hyphantria cunea, larvae (yellow, brown, and black, very 
 hairy). 
 
 Both of the insects this captive thrasher seemed to have (difficulty 
 in keeping down are eaten in the wild state. All of the other insects 
 accepted belong to "specially protected" species. 
 
 Bluebird (Sialia sialis) : — 
 ■Accepted: 
 
 ISOPODA. 
 
 Oniscus asellus, 2. 
 Orthoptera. 
 Gryllid^. 
 
 Gryllus sp., 2. 
 Coleopteea. 
 
 CaRABIDjE. 
 
 Anisodactylus discoideus (black and brownish-yellow). 
 
 HYDROPHILIDiE. 
 
 Hydrophilus triangularis (shining greenish -black), ate 
 parts of one broken by bluejay. 
 
 Refused: 
 
 Hymenoptera. 
 ApidjE. 
 
 Apis mellifera worker (brown). 
 Spermatophyta. 
 
 AQUIFOLIACEiE. 
 
 Ilex opaca, berries (red). 
 Disregarded: 
 
 COLEOPTERA. 
 
 COCCINELLID^. 
 
 Adalia bipundata (red, yellow, and black). 
 
1912.] 
 
 NATURAL SCIENCES OF PHILADELPHIA. 
 
 351 
 
 CHRYSOMELID.ffi . 
 
 Galerucella luteola (yellow and black), 15 disregarded 
 three times. 
 Hymenoptera. 
 Myrmicid^. 
 
 Tetramorium ccespitum. 
 
 Accepted and Disregarded: 
 
 COLEOPTERA. 
 
 SCARAByBID^. 
 
 Lachnosterna sp., A D. 
 Araneida. 
 
 Spider, A 3 D. 
 
 Disregarded and Refused: 
 coleoptera. 
 Carabid^. 
 
 Scarites subterraneus (black), D R 3. 
 
 Summary: Sialia sialis. — 
 
 
 Accepted. 
 
 Rejected. 
 
 Disregarded. 
 
 
 Spe- Speci- 
 cies. \ mens. 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 Spe- 
 cies. 
 
 Speci- 
 mens. 
 
 "Warningly colored" species 
 
 Others "specially defended" 
 
 1 10 
 1 
 
 
 
 1 
 
 
 2 
 1 
 
 
 16 
 1 
 
 "Non-protected" species 
 
 3 5 
 
 
 
 
 
 
 Accepted and 
 disregarded. 
 
 Disregarded and 
 rejected. 
 
 
 Spe- 
 cies. 
 
 Specimens. 
 
 Spe- 
 cies. 
 
 Specimens. 
 
 "Warningly colored" species 
 
 
 1 
 1 
 
 AO 
 3 
 
 1 
 
 DO 
 
 1 
 1 
 
 
 
 1 
 
 
 DO 
 1 
 
 
 RO 
 
 Others "specially defended" 
 
 3 
 
 "Non-protected" species 
 
 
 
 
 
 This bluebird accepted one insect — Hydrophilus triangularis — 
 which very probably is never taken by wild bluebirds. The ground 
 beetle — Scarites — disregarded and thrice refused by the captive 
 bluebird has been found in collected stomachs of the species. The 
 same is true of the only berry offered it, that of Ilex opaca, which the 
 caged bird rejected. The honey bee, which was refused, and spiders 
 and May-beetles, which were disregarded as well as accepted, are also 
 eaten by wild bluebirds. 
 
352 PROCEEDINGS OF THE ACADEMY OF 
 
 General Summary: All species of birds.'' — 
 
 [June, 
 
 " Wamingly colored ' ' species 
 Others "specially defended" 
 ' ' Non-protected "species 
 
 Ac- 
 cepted 
 
 137+ 
 97 
 152 
 
 Re- Disre- Accepted and 
 jected garded rejected, 
 
 30 23 
 
 16+ I 3 
 15+ 1 
 
 A 31 
 10 
 33. 
 
 Accepted and 
 disregarded. 
 
 R10+ Al 
 
 5 4 
 
 16 4 
 
 Dl 
 
 2 + 
 3 
 
 Accepted, rejected, 
 and disregarded. 
 
 Disregarded 
 and rejected. 
 
 "Wamingly colored" species.. 
 Others "specially defended"... 
 ' ' Non-protected " species 
 
 ^0 
 
 RO 
 
 
 
 
 
 24+ 
 
 4 
 
 DO 
 
 
 5 
 
 D3 
 1 
 
 
 R3 
 3 
 
 
 Totals.— 
 
 Ac- 
 cepted. 
 
 "Wamingly colored" species .... ■ 169+ 
 
 Others "specially defended" * 111 
 
 "Non-protected" species i 213+ 
 
 Percent- 1 Percent- 
 age age dis- 
 rejectedJregarded. 
 
 17.99 ' 11.29 
 17.02 I 4.25 
 18.97 i 3.28 
 
 It appears from this final summary that Judd's captive birds 
 rejected practically the same proportion of the "non-protected" 
 species offered them as of the "specially protected" group. The 
 result, therefore, is quite different from that reached in some other 
 series of experiments. Although it harmonizes with what we believe 
 is the average influence of predaceous animals, as a whole, upon their 
 prey, i.e., an indiscriminate one, it is no more worthy of respect 
 than other experimental results, for the behavior of the animals 
 experimented with did not correspond with the natural habits of 
 their species. This is amply shown by the cases (a few not noted) 
 in which the birds would not eat articles of food that have been 
 found in the stomachs of wild individuals of the same species. 
 This is more than 38 per cent, of all the items (not specimens) 
 offered that were either disregarded or rejected; if so large a pro- 
 portion of the experimental results are manifestly untrustworthy, 
 the only safe course is to place reliance in none of them. 
 
 " Includes Buteo platypterus, Archilochus colubris, Passer domesticus, Junco 
 hy emails, Zonotrichia alMcollis, and Toxostoma rufum, for which no summaries 
 were presented in previous pages. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 35S 
 
 Bibliography of Other Experiments in America. 
 
 An annotated bibliography will sufficiently illustrate the character 
 of other American experiments upon the food of birds. Few of 
 them have any special reference to the efficiency of protective 
 adaptations. The bibliography does not include citations to papers 
 on aviculture nor on the winter feeding of birds. These are very 
 numerous and their only merit from our present standpoint is that 
 they afford much proof, if proof of the obvious were needed, that 
 birds, both free and confined, will readily accept foods with which 
 their species has never had experience under natural conditions. 
 
 BoLLES, Frank. Young Sapsuckers in Captivity. Auk, IX, No. 2,. 
 
 April, 1892, pp. 109-119. 
 Proof that they can live a long time on a diet of syrup with very 
 
 few insects. 
 Carpenter, F. H. Screech Owls Breeding in Confinement. Ornith. 
 
 andOologist, 8, No. 12, December, 1883, pp. 93, 94. 
 "I fed them exclusively on frogs They seemed to prefer 
 
 them to any other food, which led me to believe that they 
 
 constituted no mean portion of their regular fare when at 
 
 liberty." 
 This inference is not supported by the results of stomach examina- 
 tions. Dr. Fisher found frogs in only two out of a total of 255- 
 
 stomachs examined {Bui. 3, U. S. Biological Survey, 1893, pp. 
 
 169-173). 
 Collins, C. W. Some Results from Feeding Eggs of Porthetria 
 
 dispar to Birds. Journ. Economic Ent., 3, No. 4, August, 1910, 
 
 pp. 343-346. 
 Some English sparrows and a pigeon were tested. In all cases 
 
 it was necessary to force the birds to eat the eggs. Eggs in 
 
 dough fed to English sparrow were mostly rejected. 
 Collins, J. W., et al. [Food of Young Ruffed Grouse.] Rep. 
 
 Comm. Inland Fisheries and Gaine, Mass., 1900 (1901), p. 43. 
 Some young ruffed grouse which had been fed on maggots, lettuce, 
 
 and young clover were given grain, and as a result died. Were 
 
 they also given gravel? 
 F. G. The Pine Grosbeak in Confinement. Ornith. and Oologist, 
 
 9, No. 4, April, 1884, p. 41. 
 Fond of corn meal and milk, apple seeds, beechnuts, and buds 
 
 and seeds of pine and spruce. 
 FoRBUSH, E. H. [Food consumed by two young crows.] Useful 
 
 Birds and Their Protection. Mass. Bd. Agr., Boston [1907], 
 
 pp. 45-48. 
 Chiefly concerns the quantity of food. Toads, frogs, and sala- 
 manders, often stated to be "distasteful," were eaten. 
 FoRBUSH, E. H., and Fernald, C. H. The Gypsy Moth, Porthetria 
 
 dispar (Linn.). Mass. State Bd. Agr., 1896. 
 
354 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 On pp. 231 and 239 it is stated that gypsy moth eggs were fed to a 
 confined EngUsh sparrow and a crow. The former ate them 
 voluntarily, but "did not appear to relish them"; the- latter 
 would not take them except when they were concealed within 
 other food. 
 
 It will be noted that in Collins' experiment (see above) the English 
 sparrow took the eggs only when they were forced upon it. 
 Herbick, F. H. The Home Life of Wild Birds. New York, 1901. 
 
 Young kingfishers rejected raw meat, but throve on fish in cap- 
 tivity (p. 92). 
 Hodge, C. F. Our Common Birds. Nature Study Leaflet, Biol. 
 Ser. No. 2, Worcester, Mass., 1899. 
 
 A young cedarbird took flies, poke berries, cabbage worms, "edema" 
 larvse, ants, fall web worms (a little sparingly), bush cranberries, 
 and peppermint drops (p. 15). Mockingbirds accepted meal- 
 worms and spiders (p. 19). 
 Hodge, C. F. [Food of Young Ruffed Grouse.] Rep. Comm. 
 Fisheries and Game, Mass., 1903 (1904), pp. 182, 183. 
 
 "I tested them with a great variety of prepared foods — grated 
 egg, bread cruitibs, scraped raw meat, grated boiled meat, grits, 
 boiled rice, millet and other small seeds, grass, clover, chickweed, 
 partridge, and wintergreen berries, etc. They would either 
 pay no attention to any of these things, or, if they did pick at 
 them at all, would not do so but once." Foods accepted were 
 sweet curds, earthworms, mosquito larvae, plant lice, mealy 
 bugs, thrips, mealworms and maggots. 
 Hodge, C. F. A Summer with the Bluebirds. Bird Lore, 6, No. 2, 
 March-April, 1904. 
 
 "In my series of feeding tests I brought in a number of potato 
 beetles and thoughtlessly dropped a large larva into an open 
 mouth, before observing whether they would take them of their 
 own accord. I notieed that they picked them up once apiece, 
 wiped their bills in disgust, and declined to touch them again. 
 Next morning one of the birds was dead under the perch" (p. 45). 
 Hodge, C. F. [Food of Young Ruffed Grouse.] Rep. Comm. 
 Fisheries and Game, Mass., 1904 (1905), pp. 132, 133. 
 
 Gives a long list of foods accepted; pears and peaches were scarcely 
 more than tasted; thorn-apples, barberries, and black alder 
 berries were not refused, but were taken in large quantities; 
 they took quantities of all sorts of leaves except grape, snowball, 
 artichoke, and Rosa rugosa. 
 
 Thorn-apples and black alder berries are commonly eaten by wild 
 ruffed grouse. See Biological Survey, Bui. 2^, 1905, pp. 36-38. 
 Hodge, C. F. [Food of Ruffed Grouse in Confinement.] Rep. 
 Comm. Fisheries and Game, Mass., 1905 (1906), pp. 65-68. 
 
 Gives names of numerous food items accepted. Berries of black 
 alder were taken sparingly; oats and barley were eaten spar- 
 ingly; peas and beans were refused. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 355 
 
 Hodge, C. F. [Food of Young Ruffed Grouse.] Rep. Comm. 
 Fisheries and Game, Mass., 1907 (1908), p. 70. 
 
 Two died from swallowing objects too large to pass into gizzard 
 (black cricket and large spider). This certainly was not the 
 cause of death. A young ruffed grouse's digestive apparatus 
 would quickly dispose of two such soft-bodied insects. 
 Hodge, C. F. [Report .... relative to the Propagation of Ruffed 
 Grouse and Quail in Confinement.] Rep. Comm. Fisheries and 
 Game, Mass., 1908 (1909), pp. 60-69. 
 
 On pp. 60 and 61, Hodge says: "I encountered a new difficulty 
 against which we must be on our guard in the future. Striped 
 plant bugs were abundant on the grass, and were easily obtained 
 by sweeping with insect nets. The young chicks [of ruffed 
 grouse] ate them greedily, and simply went to sleep and died as 
 if they had been chloroformed. These bugs had the strong 
 odor of squash bugs, by feeding which to toads Conradi found 
 that they died as though they had been poisoned with chloro- 
 form." 
 
 "Conradi found that five or six squash bugs might be sufficient 
 to kill a toad, and Miss Morse has fed as many as eleven to a 
 bobwhite at a single meal. Plant bugs are not so strong as 
 squash bugs, and I have observed a toad eat over 250 of them in 
 a day without showiiig ill effects. Still, while this evidence is 
 not conclusive, .... I think that we should be more careful in 
 future not to feed too many strong-smelling bugs to young 
 grouse chicks." 
 
 Dr. Hodge's experience with the young grouse, and the bluebird, 
 above noted, being killed by eating certain insects, is unsupported 
 by other testimony, and the observations leading to his con- 
 clusions are not scientifically exact. 
 
 The reference to Conradi's experiments is incorrect. The toads 
 when confined in small bottles were killed by the vaporized 
 secretions of squash bugs; they were not killed by eating the 
 bugs. The feeding of bobwhites is described on pp. 64-67. He 
 justly remarks: "The most careful artificial feeding of a flock in 
 confinement cannot approach in variety the food of wild birds" 
 (p. 64). Reports of the Massachusetts Commissioners on 
 Fisheries and Game for other years contain notes on the feeding 
 of game birds in captivity, but not in relation to "protected" 
 insects. 
 Hyslop, J. A. The False Wireworms of the Pacific Northwest. 
 Bui. 95, U. S. Bur. Ent., Part V, 1912. 
 
 In the discussion of natural enemies (p. 84) are reports on experi- 
 mental feedings of adult Eleodes chiefly to various gallinaceous 
 birds. Chickens, ducks, the Reeves pheasant, and silver pheas- 
 ant ate the beetles, while turkeys refused them, and golden and 
 Lady Amherst pheasants would not notice them. The author 
 says,. "However, these birds seemed quite annoyed by our 
 presence and might have eaten the beetles had they not been 
 frightened." 
 
356 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 Jenkins, W. E. [Blue Jay in Confinement.] Ornith. and Oologist, 
 9, No. 3, :March, 1884, p. 36. 
 Principal foods are meat, beechnuts, and corn. 
 Nash, C. W. The Birds of Ontario in relation to Agriculture. 
 Ont. Dep. Agr., Toronto, 1901. 
 On p. 44 are the results, as to quantity of cutworms and earth- 
 worms, experimentally fed to a young robin. 
 Owen, D. E. Notes on a Captive Hermit Thrush. Auk, XIV, 
 No. 1, January, 1897, pp. 1-8. 
 Notes on quantity of earthworms and beef eaten. Worms from 
 manure hill refused, those from garden eaten. 
 Peckham, E. G. [Fowls Eating Argiope riparia]. Occas. Papers, 
 Nat. Hist. Soc. Wis., I, 1889, p. 72. 
 This deep black and brilliant yellow spider seems to lack "one 
 means of defence common among conspicuous creatures, i.e., 
 the possession of a nauseous flavor." "Some chickens, to 
 which she was offered, ate her with relish. " 
 Reiff, W. Some Experiments on the resistance of Gypsy Moth 
 Eggs to the Digestive Fluids of Birds. Psyche, 17, No. 4, 
 August, 1910, pp. 161-164. 
 Eggs concealed in other food were fed to a German canary, a 
 chaffinch, a yellow hammer, a Japanese robin, a screech owl, 
 and a carrier pigeon. The eggs given to the first three birds 
 were put in pieces of bread. In each case part of them were 
 picked out and rejected. 
 For a more complete review of the various experiments in feeding 
 gypsy moth eggs to birds, see Auk, 28, No. 2, April, 1911, pp. 
 285, 286. 
 Scott, W. L. Baltimore Oriole. Ornith. and Oologist, 8, No. 11, 
 November, 1883, p. 86. 
 "He is particularly fond of hard-boiled egg, bread, and finely 
 chopped meat." 
 Stickney, J. H., and Hoffmann, R. Bird World, Boston, 1898. 
 An unconfined yellow-throated vireo took cankerworms and 
 many black ants (pp. 106-112). 
 Treadwell, D. [The Food of Young Robins.] Proc. Boston Soc. 
 Nat. Hist., VI, 1859, pp. 396-399. 
 Discusses amount of earthworms and beef eaten per day. 
 Weed, C. M., and Dearborn, N. [Food of a Captive Crow.] 
 Birds in Their Relations to Man, 1903, pp. 61, 62. 
 On quantity of fish consumed. 
 Wheelock, I. G. Nesthngs of Forest and Marsh, Chicago, 1902. 
 Young bluebirds were fed yolk of hard-boiled eggs, cracker crumbs, 
 and earthworms (p. 34). 
 
 Summary. 
 
 From the writer's point of view, three main conclusions regarding 
 the experimental tests of the efficiency of protective adaptations 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 357 
 
 against natural enemies are unavoidable: (1) The experiments 
 are very inconsistent ; (2) They have been misinterpreted, and (3) 
 They are not trustworthy guides to behavior under natural con- 
 ditions. 
 
 The Experiments are very Inconsistent. — Inconsistency in the details 
 of various series of experiments have been set forth in previous 
 pages (see pp. 298, 300, 313, 316 and 319). Inconsistency 
 in the results of entire series is plainly shown by the strongly 
 contradictory conclusions different experimenters have drawn. 
 Thus Weir, Poulton, Marshall, Pocock, and Finn, for instance, thought 
 their experiments supported the selectionist theories concerning 
 protective adaptations, while Butler, Manders, Punnett, Plateau, 
 Reighard, and Pritchett, among others, drew just the opposite 
 conclusion. Beddard's opinion was that distastefulness was not 
 more definitely associated with conspicuous colors, than with plain 
 ones. The characteristic inconsistency of experimental results are 
 described by him in the following language:'^ "None of these 
 experiments are thoroughly satisfactory; it is so difficult to interpret 
 them, and they are often contradictory, for a bird will eat one ,day 
 what it has refused before. The experiments that have been 
 made are like most other statistics — they may be made to prove 
 anything." 
 
 The Experiments have been Misinterpreted. — This charge weighs 
 not so much against the experiments themselves as against their 
 makers, but it throws doubt upon the desirability of such tests, 
 since the personal equation is so large a factor in the interpretation 
 of results. 
 
 Definite instances of misinterpretation have been cited in 
 previous pages (295, 303, 305-316, 325 and 328-330). A 
 chronic case is well illustrated by the following quotations from 
 Prof. E. B. Poulton (Trans. Ent. Soc. Lond., 1902): 
 
 "A mantis is probably less affected in this respect [food preferences] 
 by confinement . than a vertebrate animal; but the same general 
 criticism will probably hold in both cases — that while the rejection 
 of an insect by a not over-fed insectivorous animal in captivity is 
 evidence of unpalatability or dislike, its acceptance is not sufficient 
 evidence of appreciation or that it constitutes an element of the 
 normal diet. An insect may be eaten readily in captivity which- 
 would be rejected or only eaten under the stress of hunger in the 
 wild state" (p. 317). 
 
 « Animal Coloration, 1892, p. 166. 
 
358 PKOCEEDINGS OF THE ACADEMY OF [June, 
 
 "It has already been pointed out that the acceptance of insects 
 by insectivorous animals in captivity is no proof of their normal 
 likes or dislikes in a wild state. Such acceptance only proves what 
 their action would be when they had been, from some exceptional 
 cause, kept without their normal food in its usual quantity and 
 variety. Hence the fact that Acraeas were devoured [by a ground 
 hornbill] is no evidence that they are normally eaten except in a 
 time of unusual hunger. On the other hand, the rejection of two 
 L. chrysippus, after three Acrseas had been readily eaten, indicate 
 that the former butterfly is decidedly distasteful to this species of 
 
 bird" (p. 34-8) 
 
 "Byblia ilithyia was .... distinguished [by baboons] from an 
 .Acrcea, but this by no means proves that the resemblance is not 
 
 beneficial under natural conditions (p. 388) Considering 
 
 what has been already argued about insect-eating animals in con- 
 finement, the acceptances (excluding the Hesperiidae) probably do 
 not justify the conclusion that the Lepidoptera were palatable, or 
 that they would be sought for in the wild state except under the 
 stress of hunger" (p. 389). 
 
 "It has already been pointed out that the refusal or evident 
 dislike of insect food by captive animals is trustworthy evidence of 
 unpalatability, while acceptance is not proof of palatability " (p. 436). 
 It is self-evident that this oft-repeated dictum is merely special 
 pleading for the admission of as much as possible of the evidence 
 favorable to the theories, and the exclusion of as great a proportion 
 as possible of the evidence that might be unfavorable. So plain is 
 this fact that even Mr. G. A. K. Marshall, collaborator with Prof. 
 Poulton in the paper quoted from, severely criticized the Professor's 
 attitude. He says'' in part: 
 
 "There is. too emphatic an insistence upon the possibility of error 
 where an insect is accepted; for it practically casts suspicion upon 
 every such case. On the other hand, the possibility of error in the 
 other direction is not indicated. " 
 
 The Experiments are not Trustworthy Guides to Behavior under Natural 
 Conditions. — The writer is by no means the first to question the 
 analogy of behavior under experimental to that under natural 
 conditions. The idea is put briefly by L. W. Kline in an article on 
 " Methods in Animal Psychology" :'* "Nothing so shrinks and in- 
 
 «3 Trans. Eni. Soc. Lond., 1908, p. 140. 
 ^Amer. Journ. of Psychol, 10, 1898-9, p. 276. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 359 
 
 hibits completely the fulness and variety of an organism's activities 
 than prison life and fear." 
 
 In groups as low even as the Amphibia behavior in confinement 
 is far from natural. Prof. C. O. Whitman found that Nedurus 
 ordinarily refused food in captivity on account of its extreme timid- 
 ity. He says:'° "The first adults which I kept in captivity in a 
 large aquarium refused to eat pieces of raw beef or small fish, whether 
 dead or alive. For months they went on, seeming entirely indifferent 
 to any proffered food, not paying the least attention, so far as I 
 noticed, to tempting morsels dropped quietly in front of them or 
 held in suspension before them. Living earthworms and insect 
 larvse were presented to them, all of which were known to be palatable 
 to the creature in its natural habitat; but nothing availed to draw 
 attention or elicit any evidence of hunger. Quiet and wholly indif- 
 ferent in outward behavior, yet the animals were actually starving 
 or wasting away." 
 
 Many snakes will not take food in captivity, and it is therefore 
 necessary to force food down their throats to prevent death from 
 starvation. Captivity greatly modifies the behavior of some other 
 reptiles also, as is well stated in the following quotation from H. H. 
 Newman: "In order to understand an animal one must live with 
 it, must spend long hours, quiet days, in thoughtful observation of 
 it, as it pursues its daily round of occupations. This I have had an 
 opportunity of doing, and I now feel that I have a really personal 
 acquaintance with at least five species of tortoises 
 
 "Studies of this sort should, I believe, precede experimental 
 studies, for sometimes shyness or weariness might be mistaken for 
 stupidity, and sullenness for sluggishness in reaction. As a rule, 
 the more highly organized and alert species of tortoises display, 
 when in captivity, the greatest degree of sullenness, and hence their 
 actions in confinement very poorly represent their true character. 
 The species, on the other hand, that are less highly organized are 
 
 the species that act more nearly normally when in captivity 
 
 Captivity inhibits normal activity in nearly all tortoises; conse- 
 quently I abandoned at an early stage of my work the observation 
 of specimens in confinement and devoted myself to long-continued, 
 and at times tedious, observation of the various species as they live 
 in their active environment 
 
 "Extreme sullenness characterizes the behavior of Aspidonedes 
 
 95 Biol. Lectures, 1898 (1899), pp. 295, 296. 
 
:360 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 while in confinement. If kept in a room they hide behind furniture 
 .and remain motionless for hours and almost days. When put in 
 aquatic enclosures they immediately bury themselves in the mud 
 and seem to remain there for months. Nothing will induce them to 
 ■eat or to take any interest in their surroundings. If caught while 
 making their nest, they are sometimes forced to lay the eggs, but 
 never make a nest in confinement. The eggs are simply dropped 
 about on land or in the water, and are usually crushed when found. 
 None of their normal characteristics are in evidence, and it would be 
 .a waste of time to attempt to draw conclusions about their disposi- 
 tion or intelligence from their actions in captivity. "'* 
 
 Prof. Charles W. Hargitt makes a similar but more general criticism 
 •of the experimental method of studying animal behavior, as follow.? : 
 "I have made the field work emphatic whenever at all practicable. 
 I have elsewhere'' emphasized the crying need for larger attention 
 to this phase of experimental work, believing that in many cases 
 it is all but impossible to secure trustworthy results as to behavior 
 of animals where the work has been done under such unusual, un- 
 natural, and artificial conditions as most laboratory provisions afford. 
 
 "What right has one to assume that the actions of an animal taken 
 rudely from its natural habitat and as rudely imprisoned in some 
 improvised cage are in any scientific sense an expression of its normal 
 behavior, either physical or psychical? Is it within the range of the 
 calculus of probability that conclusions drawn from observations 
 made upon an animal in the shallow confines of a finger-bowl, but 
 whose habitat has been the open sea, are wholly trustworthy? It is 
 no part of my purpose to discredit the laboratory or laboratory 
 appliances as related to such investigations. They are indispensable. 
 But at the same time let it be recognized that they are at best but 
 artificial makeshifts whose values, unless checked up by constant 
 appeal to nature, must be taken at something of a discount. This 
 must be especially the case with higher organisms. Some of these 
 may, of course, be readily domesticated, or made more or less at 
 home in aquaria or vivaria; but not a few absolutely fret their 
 lives out, are never at ease, and probably never give expression to a 
 natural reaction under such conditions. It seems to the writer 
 until one has been able to place his specimens under conditions 
 
 ™"The Habits of Certain Tortoises," Journ. of Corn-par. Neurology and 
 Psychol., XVI, 2, March, 1906, pp. 126, 127, and 135. 
 
 " " Observations on the Behavior of Tubicolous Annelids," Journ. Exp. Zooh, 
 Vol. 7, 1909, p. 157. 
 
1912.] NATURAL SCIENCES OF PHILADELPHIA. 361 
 
 approximating the natural, or has at least brought them to a state 
 of semi-domestication, where in food taking, evidence of health, etc., 
 they are at ease, he has small right to dogmatize as to conclusions, 
 or presume to make such conclusions the basis of so-called laws of 
 animal behavior. Not a little of recent investigations along the 
 lines of animal behavior has been vitiated at just this point, and 
 must be repeated to be made trustworthy. The amazing mass of 
 contradictory results which has loaded the. literature of recent 
 years is attributable to some extent to this misfortune."'^ 
 
 With regard to experimentation with captive birds. Prof. S. A. 
 Forbes, the founder of economic ornithology, says:'' "This evi- 
 dently shows only what the bird will eat when restrained of its 
 liberty, of such food as may be placed before it, and furnishes few 
 data which we can use with safety in making up an account of its 
 food in freedom, when foraging for itself. The state of confinement 
 is so abnormal for a bird that on this account, also, we can rarely 
 reason from its habits in that state to its ordinary habits. This 
 method is, therefore, available only for the solution of a few separate 
 questions." 
 
 The assertions of these authors regarding the modifying effects 
 of captivity upon behavior apply more pertinently to no set of 
 experiments than those which have been conceived to be tests of 
 the food preferences of insectivorous animals in relation to pro- 
 tective adaptations. 
 
 The writer has asserted that the experiments are not trustworthy 
 guides to behavior under natural conditions, and he expects to prove 
 this by citing evidence along two lines, viz. : (1) Animals accept in 
 captivity articles of food which they not only do not eat in the wild 
 state, but with which their species probably has never had experience, 
 and (2) animals reject in captivity articles of food which are not only 
 occasionally eaten by wild members of the species, but which may be 
 very important elements of the subsistence as a whole. 
 
 (1) Acceptances. — This point really needs no proof. Universal 
 experience with the feeding of all kinds of captive animals confirm it. 
 The coarse brown bread (containing oats, shorts and molasses) given 
 to the bears, in some zoological parks, the chopped-up beets, carrots, 
 potatoes, etc., of which the parrots, cranes, and certain rodents are 
 fond, sufficiently illustrate foods relished in confinement by animals 
 
 98 Journ. of Animal Behavior, Vol. 2, No. 1, January-February, 1912, pp. 51, 52. 
 » Bui. III. State Lab. Nat. Hist., Vol. I, No. 3, 1880, pp. 86, 87. 
 
362 PROCEEDINGS OF THE ACADEMY OF [June, 
 
 to which they are unknown in the wild state. The experiments of 
 Pocock and Butler, resulting in the acceptance of many British 
 insects by a variety of foreign mammals and birds, illustrate the same 
 point. As noted before, the acceptance of butterflies by some of 
 Finn's birds signifies no more, concerning their natural food habits, 
 than does their acceptance of boiled rice. It means no more than 
 the eating of silver fish, clothes moths, and mealworms by Mrs. 
 Nice's bobwhites. 
 
 The point need be no further elaborated. We are forced to 
 conclude that acceptance of various items of food by captive animals 
 is no indication whatever that they are eaten by the same species 
 in the wild state. 
 
 (2) Rejections. — This point really follows from analogy the con- 
 clusion just cited. There is no logic in regarding rejections as 
 indicative of natural tastes, when acceptances are plainly shown not 
 to be. But evidence to prove the case is much harder to obtain, 
 and it is for this reason that we have been compelled to endure the 
 style of argument that asserts " refusal .... is trustworthy evidence 
 of unpalatability, while acceptance is not proof of palatability." 
 
 Fortunately, however, we have information regarding the choice 
 of food by a number of animals, both in captivity and under natural 
 conditions. We have shown that in certain of the experiments with 
 amphibia, the animals refused articles of food which they habitually 
 eat in the natural state. For instance, this is true of the refusal 
 by the common toad of the Eastern United States of millipeds 
 (Julus), squash-bugs (Anasa tristis), and potato beetles {Leptino- 
 tarsa decemlineata) . Prof. Whitman found that ordinary articles 
 of the natural diet were refused by captive Necturus. Snakes, in 
 particular, often refuse all food in confinement. Is this "trustworthy 
 evidence of unpalatability?" The writer had the care for a year of 
 six prairie rattlesnakes (Sistrurus cate?iatus). Live mice and birds 
 put in their cage were killed, but not eaten. No food was taken 
 naturally and they were kept alive only by putting meat well down 
 their gullets with long-jawed forceps. 
 
 Beddard found that a green woodpecker made great objection to 
 eating a single earwig, yet Newstead found twenty-three of these 
 insects in the stomach of a wild bird of this species. Finn found 
 that captive red-whiskered bulbuls refused Acrcea, but an observer 
 in India saw the birds feeding the "most distasteful" insect of the 
 genus to their young. So little is known regarding the natural food 
 of birds in most countries that few such comparisons can be made. 
 
1912.] 
 
 NATURAL SCIENCES OF PHILADELPHIA. 
 
 363 
 
 Fortunately, this is not the case in the United States, where we have 
 data enough to prove the point. 
 
 A selection of the more conspicuous cases of refusal of favorite 
 natural foods by the birds tested by Judd are given in the following 
 tabulation. Other instances are cited in the summaries of the 
 separate experiments (pp. 340, 346, 349, 350 and 351). 
 
 Rejected by captive birds. 
 
 A ruby-throated hummingbird 
 rejected small leaf -hoppers, flies, 
 flea-beetles, and spiders. 
 
 A bluejay refused a hen's egg, 
 living birds and mice, and mul- 
 berries. Acorns also were refused. 
 
 English sparrows rejected 
 fruiting heads of dandelion. 
 
 A snowbird, a white-throated 
 sparrow, and a song sparrow re- 
 fused seeds of lamb's quarters. 
 
 The song sparrow rej ected, also, 
 seeds of smartweed and beetles 
 of the genera Diabrotica, Hippo- 
 damia, and Lachnosterna. 
 
 Eaten by wild individuals of the 
 same species. 
 
 All of these items are com- 
 monly eaten by wild birds of 
 this species. 
 
 In a state of nature jays fre- 
 quently break up outlying nests 
 of fowls. Birds and mice are 
 preyed upon and mulberries are 
 eaten. Acorns have been found 
 in dozens of stomachs of wild 
 birds and are a very important 
 element of the normal diet. 
 
 Whole flocks of English spar- 
 rows pass days in rifling the ripe 
 involucres of this plant. 
 
 These seeds are a common 
 natural food of all three birds. 
 
 All are eaten by wild song 
 sparrows, the smartweed seeds 
 in abundance. 
 
 These facts show that the feeding reactions of various animals 
 are strikingly modified by confinement. Some animals refuse items 
 of food which are a favorite with wild individuals of the species, 
 some of which may form a notable percentage of the total subsist- 
 ence. Others refuse all food. The birds experimented upon by 
 Judd togetlier disregarded or rejected 108 articles of food. Forty- 
 two of these items have been found in stomachs of wild birds of the 
 same species that ignored or refused them in captivity. Investiga- 
 tions carried on while this paper was in preparation raised this 
 number to 42 from 35, and it must be borne in mind that subsequent 
 stomach examinations will increase, never diminish the total. The 
 
364 PROCEEDINGS OF THE ACADEMY OF [JunC, 
 
 experimental indications as to what food items are unattractive or 
 distasteful to the birds, thus, are proved to be misleading in 42 cases 
 out of 108. This makes a percentage of error of 38 (which will 
 grow larger), enough to entirely invalidate the data. Furthermore, 
 it is not probable that the data from any other series of experiments 
 are any more reliable. The conclusion cannot be avoided, therefore, 
 that the rejection of various items of food by captive animals does 
 not prove that these items are rejected by the same species tmder 
 natural conditions. 
 
 Conclusion. 
 
 It has been demonstrated that behavior of captive animals toward 
 food is not a reliable indication of what wild individuals of the same 
 species would do in the presence of the same food. In other words, 
 since the feeding habits of an animal in captivity may vary widely 
 from its known habits in the natural state, there is no avoiding the 
 conclusion that the results obtained under experimental conditions, 
 do not indicate the part the animal might play in natm-al selection. 
 
 We must conclude, therefore, since acceptances and rejections 
 in experiments bear no close relation to food preferences under natural 
 conditions, that the value of experiments to determine the efficiency 
 of warning colors, and other protective adaptations of prey, is very 
 questionable. Having no certain value in themselves, they must 
 be checked up with definite knowledge of the natural food habits. 
 This information is obtained by collecting animals with freshly 
 captured prey and by examination of pellets, castings, and the 
 contents of stomachs or other portions of the alimentary canal. 
 There is no possibility of going back of such evidence on the choice 
 of food, nor is there any need of so doing. 
 
 Since this evidence is sufficient in itself, and since experimental 
 data must be supported by it to be worthy of any consideration, why 
 perform the experiments? The same time expended in collecting 
 trustworthy data regarding the natural food habits of animals would 
 bring much greater returns, and the result would be truth, not 
 imaginative inferences from abnormal behavior.