We have next the interesting fact that like our melanic moths the dark form is replacing the "type." At the time of Ober's visit the type was already in a minority, but now it is nearly or perhaps actually extinct, though the black form is one of the commonest birds on the island. Austin Clark found no specimen when he collected there in 1903-4, though formerly it was not uncommon in the vicinity of Kingston and in the immediate windward district of St. Vincent.
The Grenadines are geographically just south of St. Vincent, though separated by a deep channel. In these islands no black forms have yet been taken, but Grenada, the next island to the south, has both normals and blacks. There are trifling differences of size between the Grenada birds and those from St. Vincent, the Grenada specimens being slightly smaller and for this reason they have received distinct names, the form marked with yellow and white being calledGodmani(Cory) and the black,Wellsi(Cory), but this merely introduces a useless complication.There is evidence that in Grenada, as in St. Vincent, the black is gradually ousting the original type, but the process has not gone so far as in St. Vincent. Austin Clark very properly compares this case of the Sugar-birds with that ofPapilio turnus, which as is well-known, has a black female in the southern parts of its distribution, in addition to a female of the yellow type, but in the Northern States the black female does not occur.
During the present year P. R. Lowe, who lately studiedCoerebason a large scale in the West Indies, has published an important paper on the subject.[29]He calls attention to the fact that Cory recently found a black form ofCoerebaon Los Roques Islands, and he himself discovered another on the Testigos Islands. Both localities are on the coast of Venezuela, far from St. Vincent and Grenada. The whole problem is thus further complicated by the fact that the black varieties have, as we are almost driven to admit, arisen independently in remote places. Improbable as this conclusion may be, it is still more difficult to regard all the black forms as derived from one source. For first, they present definite small differences from each other; and secondly we have to remember a consideration of greater importance, that the very fact that each island has its own type must be accepted as proving that the localities are effectively isolated from each other, and that migration must be a very rare event.
The rarity of such illustrative cases is, I believe, more apparent than real. It is probably due to the extreme reluctance of systematists to admit that such things can be, and of course to the almost complete absence of knowledge as to the genetic behaviour of wild animals and plants. Only in such examples as this of theCoereba, where colour constitutes the sole difference, or that of the moths which have been minutely studied by many collectors, does the significance of the facts appear. The arrangement of catalogues and collections is such that much practical difficulty of a quite unnecessary kind is introduced. For example, in this very case ofCoereba, I find the British Museum has a fine series from Grenada including 3 normals and 11 black,and also 16 blacks from St. Vincent. If the black specimens from Grenada were put with the normals which are almost certainly nothing but a recessive form of the same bird, the variation would strike the eye on even a superficial glance at the drawer. But following the notions so naively expressed in the passage quoted above from W. N. Lawrence, the blacks from Grenada are put apart together with the other blacks from St. Vincent, though two of them were shot on the same date as one of the normals.
Overlapping Forms
The facts of the distribution of local forms on the whole are consistent with the view that these forms come into existence by the sporadic appearance of varieties in a population, rather than by transformation of the population as a whole. Of such sporadically occurring varieties there are examples in great abundance, though by the nature of the case it can be but rarely that we are able to produce evidence of a previous type being actually superseded by the variety. When the two forms are found co-existing in the same area they are usually recorded as one species if intergrades are observed, and as two species if the intergrades are absent. On the other hand when two forms are found occupying separate areas, when, that is, the process of replacement is completed in one of the areas, then forthwith each is named separately either as species or subspecies. Successive observations carried out through considerable periods of time would be necessary to establish beyond question that the history proceeds in one way rather than another. Such continuity of observation has for the most part never been attempted. The kind of information wanted has indeed only been lately recognized, and really critical collecting is a thing of only the last few decades. The methods of the older collectors, who aimed at bringing together a few typical specimens of all distinct forms, are of little service in this class of inquiry, which is better promoted by the indiscriminate collection of large numbers of common forms from many localities. When this has been done on a comprehensive scale we shall be in a position to form much more confident judgments as to the general theory of evolution.
Some little work of the kind has however been done and the results are already of great value. Seeing that the differentiation of local forms is only made possible by isolation, it necessarilyhappens that the collector finds one form in one locality and another in a distinct locality, and there is no evidence as to the behaviour which the two representative species might exhibit if they came into touch with each other. In the most familiar examples of such distinction each inhabits an island, completely occupying it to the exclusion of any other similar form. It can only be when the two representative species occupy parts of a continental area connected with each other by regions habitable for the organism in question, that there is a chance of seeing the two forms in contact. Often also, even where this condition is satisfied, the habits, social organisation, or some other special cause may act as a barrier which prevents the distinguishable forms from ever coming into such complete contact as to interbreed or to behave as a genetically continuous race. When genetic continuity is ensured by a constant diffusion of the population over the whole area which they inhabit there will manifestly be no formation of local races. The practical uniformity, for example, of so many species of birds which inhabit widely extended ranges of Western Europe is doubtless maintained by such constant diffusion. When, as in the case of the Falcons, many localities have peculiar forms, the fact may be taken as conclusive evidence that there is little or no diffusion; and when we find in such a species as the Goldfinch that in spite of migratory fluctuations there are nevertheless geographical races fairly well differentiated, it may similarly be inferred that these fluctuations habitually move up and down on paths which do not intermingle. There are however a few examples of animals, not given to much irregular wandering, which occupy a wide and continuous range of diversified country and are differentiated as local races in two or more districts, though the distinct races meet in intervening areas. Of these the most notorious illustration which has been investigated with any thoroughness is that of the species ofColaptes(Woodpeckers) known in the United States as Flickers. The study of the variations of these forms, made by J. A. Allen[1]is an admirable piece of work, with whichevery student of variation and evolutionary problems should make himself familiar. The two forms with which we are most concerned are known asC. auratusandC. cafer, and are very strikingly different in appearance. In size, proportions, general pattern of colouration, habits, and notes, the two are alike, but they differ in the following seven respects as stated by Allen.
These differences are illustrated in the accompanying coloured plate, which has been most kindly prepared for me under the instructions of Dr. F. M. Chapman of the American Museum of Natural History. Before going further it is worth considering the nature of these differences a little more closely. All but the last are large differences which no one would overlook even in a hasty glance at the birds. If the only distinction lay in the colour of the quills we might feel fairly sure thatauratuswas a recessive form ofcafer, and so probably it is in this respect. Similarly the black malar stripe ofauratusis in all probability recessive to the red malar stripe ofcaferand I imagine the pigments concerned are comparable with those in the Gouldian Finch (Poephila gouldiae) of Australia. Both sexes in that species may have the head black, red, or, less often, yellow, and though it is not any longer in question that birds may breed in either plumage, I believe that the young are always black-headed and I imagine that those which become red-headed possess a dominant factor absent from the permanently black-headed birds.[2]Yellow as a recessive form of a red is certainly very common, but red and black as variants of the same pigment are less usual. In the Gouldian Finch we seem to have a case where a pigment can assume all three forms. It would be interesting to know whether the red of the malar stripes inColaptesis a pigment of the same nature as the red of the quills. Both inColaptesand inPoephila gouldiaeI have seen specimens intermediate between the black and the red, and the appearance of the part affected was exactly alike in the two cases, red feathers coming up among the black ones, and many feathers containing both red and black pigments mixed together.
The development of the scarlet nuchal crescent inauratusand the absence of this conspicuous mark incaferconstitute from the physiological point of view the most remarkable pair of differences. When the red crescent is not formed, the feathers which would bear it are exactly like the rest, and no special pigment is visible in them which one can regard as ready to be modified into red. If the crescent is due to a factor it must therefore be supposed that this factor has the power of modifying the pigment of the neck in one special place alone. Dr. W. D. Miller called my attention to the fact that a similar variation occurs in another American woodpecker, the Sapsucker,Sphyropicus varius.[3]
I do not suggest that such variations are without parallel: indeed inP. gouldiaethe factor which turns the black of the head into scarlet affects one special region of the black only, being sharply distinct from the unmodified black of the throat. These regions of the head are however often the seat of special colours in birds.[4]So also may be instanced the variety of the CommonGuillemot (Uria troile) which has a white line round the eyes and at the sides of the head where the normal has no such mark; but this line is formed in a very special place, the groove joining the eye to the ear, whereas the feathers of the nuchal crescent are not ostensibly distinguished from those adjacent.[5]
The transposition of the brown and the grey on the back and front of the neck also constitutes a very remarkable difference. If either grey or brown depends on a factor then it must be supposed thatauratushas one of these factors andcaferthe other.
From these several considerations it is quite clear that ifauratusandcaferare modifications of the same type produced by presence or absence of factors, several independent elements must be concerned, and to unravel their inter-relations would be most difficult even if it were possible to breed the types under observation, which is of course quite beyond present possibilities.
The distribution of the two is as follows. On the east side of the ContinentC. auratus, relatively pure, occupies the whole of Canada and the States from the North to Galveston. Westward it extends across the whole continent in the more northern region to Alaska, but in its pure form it only reaches down the Pacific coast to about the northern border of British Columbia. Its southern and western limit is thus roughly a line drawn from north of Vancouver, southeast to North Dakota and then south to Galveston.C. caferin the comparatively pure form inhabits Mexico, Arizona, California (except Lower California and the opposite coast), central and western Nevada, Utah, Oregon, and is bounded on the east by a line drawn from the Pacific south of Washington, south and eastward through Colorado to the mouthof the Rio Grande or the Gulf of Mexico. Between the two lines thus roughly defined is a band of country about 1,200-1,300 miles long and 300-400 miles wide, which contains some normal birds of each type, but chiefly birds exhibiting the characters of both, mixed together in various and irregular ways. Even in the areas occupied by the pure forms occasional birds are recorded with more or less indication of characteristics of the other form, but within the area in which the two forms are conterminous, the mixed birds are in the majority. The condition of these birds of mixed character is described by Allen as follows:
"As has been long known—indeed, as shown by Baird in 1858—the 'intermediates' or 'hybrids' present ever-varying combinations of the characters of the two birds, from individuals ofC. auratuspresenting only the slightest traces of the characters ofC. cafer, or, conversely—individuals ofC. caferpresenting only the slightest traces of the characters ofC. auratus—to birds in which the characters of the two are about equally blended. Thus we may haveC. auratuswith merely a few red feathers in the black malar stripe, or with the quills merely slightly flushed with orange, orC. caferwith either merely a few black feathers in the red malar stripe, or a few red feathers at the sides of the nape, or an incipient, barely traceable scarlet nuchal crescent. Where the blending of the characters is more strongly marked, the quills may be orange-yellow or orange-red, or of any shade between yellow and red, with the other features of the two birds about equally blended. But such examples are exceptional, an unsymmetrical blending being the rule, the two sides of the same bird being often unlike. The quills of the tail, for example, may be part red and part yellow, the number of yellow or red feathers varying in different individuals, and very often in the opposite sides of the tail in the same bird. The same irregularity occurs also, but apparently less frequently, in the quills of the wings. In such cases the quills may be mostly yellow with a few red or orange quills intermixed, or red with a similar mixture of yellow. A bird may have the general colouration of truecafercombined with a well-developed nuchal crescent, or nearly pureauratuswith the red malar stripes of acafer.Sometimes the body plumage is that ofC. auratuswith the head nearly as in purecafer, or exactly the reverse may occur. Or we may have the general plumage as incaferwith the throat and crown as inauratus, and the malar stripe either red or black, or mixed red and black, and so on in almost endless variations, it being rare to find, even in birds of the same nest, two individuals alike in all their features of colouration. Usually the first trace ofcaferseen inauratusmanifests itself as a mixture of red in the black malar stripe, either as a few red feathers, or as a tipping of the black feathers with red, or with merely the basal portion of the feathers red. Sometimes, however, there is a mixture of orange or reddish quills, while the malar stripe remains normal. InC. caferthe traces ofauratusare usually shown by a tendency to an incipient nuchal crescent, represented often by merely a few red-tipped feathers on the sides of the nape; at other times by a slight mixture of black in the red malar stripe."
Such a state of things accords very imperfectly with expectations under any received theory of Evolution. As in some of the instances discussed in the first chapter we have here two fairly definite forms, nearly allied, which on any evolutionary hypothesis must have been evolved either the one from the other, or both from a third form at a time not very remote from the present, as time must be measured in evolution. Yet though intermediates exist in some quantity, no one can for a moment suggest that they are that definite intermediate from whichauratusandcaferdescend in common. One cannot imagine that the immediate ancestor of these birds was a mosaic, made up of asymmetrical patches of each sort: but that is what many of the intermediates are. It is not much easier to suppose the ancestor to have been a nondescript, with a compromise between the developed characters of each, with quills buff, malar stripes neither black nor red, with a trace of nuchal crescent, and so on. Such Frankenstein-monsters have played, a considerable part in the imaginations of evolutionary philosophers, but if it were true that there was once a population of these monsters capable of successful existence, surely they should now be found as a population occupying the neutral zone between the two modernforms. Yet, though much remains to be done in clearing up the facts, one thing is certain, namely that the neutral zone has not a definite and normally intermediate population, but on the contrary it is peopled by fragments of the two definite types and miscellaneous mongrels between them.
On the other hand, one cannot readily suppose that either form was the parent of the other. The process must have involved both addition and loss of factors, for whatever hypothesis be adopted, such changes must be supposed to have occurred. A careful statistical tabulation of the way in which the characters are distributed in the population of the mixed zone would be of great value, and till that has been done there is little that can be said with certainty as to the genetics of these characters. In the collection of Dr. Bishop of New Haven I was very kindly allowed to examine a sample, all taken at random, near together, in Saskatchewan. There were females 4 adult, 2 young; males 4 adult and 5 young. This number, though of course insufficient, is enough to give some guide as to the degree of definiteness which the characters generally show in their variations. Of the 15 birds, 8 had simply yellow quills; 2 had red; 1 was almost red but had one yellow tail-quill; 3 were intermediate and 1 was buff. As regards the malar patch, which can only be determined properly in the adult males, 1 was red, 1 was approximately red, 2 intermediate. As to nuchal crescent 4 females had none, 2 females very slight; 7 males had it, 1 had only a slight crescent, and 1 had none. In point of quills therefore 10 were definite out of 15; in point of crescent, 11 were definite out of 15; and in point of malar patch 1 only was definite out of 4. The last is a feature directly dependent on age and so counts for less, but as regards the other two features there is some indication that the factors show definiteness in their behaviour. It must be remembered that we have no knowledge what the heterozygous form may be, and in the case of red and yellow it is probably a reddish buff. The patch-works are no doubt to be compared with other well-known pied forms, and in these we must suppose the active factor broken up, which it probably can be very easily. The asymmetry, which Allen notices as so marked a feature, in thedistribution of the red and yellow quills of the tail especially, recalls that of the black markings in the pied Canaries. As is well known to students of variationssomepigment-factors insomeanimals are apparently uncontrolled by symmetry, while in other specific cases symmetry is the rule. On the other hand the blackness or redness of the malar patches is, I think, as a rule nearly symmetrical. It should be mentioned that two of Dr. Bishop's young birds belonged to the same nest, one a female withredquills, the other a male withyellow. Both are without crescent.
As to the question whether certain combinations of characters occur with special frequency, the evidence is insufficient to give a definite answer. Among all the birds I have seen in America or in England I have not yet found one having the malar patches black without any nuchal crescent. Of Dr. Bishop's 8 adults not one, however, showed the combination of the three chief features normal forauratusor forcafer.
Besides the two forms that we have hitherto considered, several other local types exist, and these throw some further light on the problem. Of these the most important in this connexion ischrysoides, which inhabits the whole of southern California and the mainland opposite. This remarkable form is as Allen says, very different fromauratusexcept that it has the quills yellow likeauratus, not red likecafer. So that we find here in the extreme west of the whole distribution a type agreeing in one of its chief features with the eastern type. Between this andcaferintergrades have, according to Allen, not been found. The relations of thischrysoidesare, Allen thinks, rather withmexicanoides, a southern, smaller race with colours more intense, which inhabits Guatemala, but however that may be, it must be regarded as acaferwhich has lost its red quills. The island of Guadeloupe off Lower California has an island form. Beyond the other side of the continent there is also an island form ofauratus, inhabiting Cuba, so that clearly the yellow quills can extend into the tropics.
The above account is in many respects incomplete, but it suffices to give an outline of the chief facts. The whole problemis complicated by the undoubted effects of an uncertain amount of migration, and in many, perhaps all, districts, the winter population differs from the summer population of the same localities. The existence of these seasonal ebbs and flows is now well known to ornithologists, and most of the bird species of temperate regions are subject to them.
Difficult as it may be to conceive the actual process of origin of the two typesauratusandcafer, it is I think still harder to suggest any possible circumstance which can have determined their development as distinct races, or which can maintain that distinctness when created. Some will no doubt be disposed to appeal once more to our ignorance and suggest that if we only knew more we should see that the yellow quills, the black "moustache" and the red crescent, specially qualifyauratusfor the north and eastern region, and the red quills, red "moustache" and absence of crescent fitcaferto the conditions of its homes. Each can judge for himself, but my own view is that this is a vain delusion, and that to cherish it merely blunts the receptivity of the mind, which if unoccupied with such fancies would be more ready to perceive the truth when at last it shall appear. Think of the range of conditions prevailing in the country occupied byauratus—a triangle with its apex in Florida and its base the whole Arctic region of North America. Is it seriously suggested that there is some element common to the "conditions" of such an area which demands a nuchal crescent in the Flickers, though the birds of thecaferarea, almost equally varied, can dispense with the same character? Curiously enough, the geographical variation ofSphyropicus varius, another though a very different Woodpecker[6]shows that conversely the nuchal crescent can be dispensed with in the Eastern form though it is assumed by the Western.[7]
Allen points out the interesting additional fact that superposed upon each of the two distinct forms,auratusandcafer, are many geographical variations which can very naturally be regarded as climatic. Each decreases in size from the North southward, as so many species do.[8]They become paler in the arid plains, and show the ordinary phases which are seen in other birds having the same distribution. Such differences we may well suppose to be determined directly or indirectly, by environment, and we may anticipate with fuller knowledge it will be possible to distinguish variations of this nature as in the broad sense environmental, from the larger differences separating the two main types ofColaptes, which I surmise are altogether independent of such influences.
It is generally supposed that phenomena like those now so well established in the case ofColaptesare very exceptional, and as has already been stated a number of circumstances must combine in order that they may be produced. I suspect however that the examples are more numerous than is commonly thought. In all likelihood the three formsSphyropicus varius,nuchalisandruberare in a very similar condition though the details have not, so far as I know, been worked out. A complex example which is closely parallel to the case ofColapteswas described by F. M. Chapman[9]at the same date as Allen's work. This is the case ofQuiscalus, the Grackles, which in the North American Continent have three fairly distinct forms which Chapman speaks of asQ. aeneus,Q. quiscula, andQ. quiscula aglaeus. The birds are all, so far as pigment is concerned, dark blackish brown, but the head and mantle have superposed a metallic sheen of interference-colours which in the various forms take different tints,bluish green, bronze green, or bronze purple. The details are complicated and difficult to appreciate without actual specimens, but the two common types are sufficiently distinct. The birds inhabit the whole area east of the Rockies,quiscula aglaeusoccupying Florida and the Southern States southwest of a band of country about a hundred miles broad extending roughly from Connecticut to the mouth of the Mississippi; andaeneustaking the area north and west of this band. In discussing this case Chapman expresses the same view as Allen does in theColaptescase, that there are two distinct populations, substantially fixed, and that the band of country in which they meet each other has a mongrel population, with no consistent type, but showing miscellaneous combinations of the character of the two chief types.
The warblers of the genusHelminthophilaprovide another illustration which has points of special interest. The two chief species areH. pinus, which has a yellow mantle and lower parts, white bars on the wings, a black patch behind the eyes and a broad black mark on the throat; andH. chrysopterawith dark grey mantle and pale whitish grey lower parts, yellow bars on the wings, and grey marks on cheeks and throat wherepinushas black. These two birds are exceeding distinct, and in addition their songs are quite unlike.H. pinusranges through the eastern United States up to Connecticut and Iowa.H. chrysopterais a northern form extending down to Connecticut and New Jersey. Both are migrants.
In these two States, where the two types overlap, certain forms have been repeatedly found which have been described as two distinct species,Lawrenceiandleucobronchialis. Dr. L. B. Bishop and Mr. Brewster showed me two long series ofHelminthophilacontaining various intergrades between the four named kinds, and details regarding these may be found in Chapman'sNorth American Warblersand in Dr. Bishop's paper in Auk, 1905, XXII. Though the characters evidently break up to some extent, the series can be represented as due to recombinations of definite factors more easily than the others which I have described. The differentiating characters are:
The grey pigment of the mantle is common to both, but is masked by the yellow inpinus, the net result being an olive-green.[10]
I am much indebted to Dr. F. M. Chapman for the loan of the coloured plate in which these distinctions are shown. It first appeared in his book,North American Warblers.
We cannot tell whetheryellowornot-yellowis due to the presence of a factor, but we may suppose that one or other gives the special colour to the parts. The black of character 3 is no doubt a dominant. Thuspinusbecomes Y1y2b andchrysopterain y1Y2B. TheLawrenceiwhich has the underpartsyellow, wing-barswhite, andblackpatches is Y1y2B andleucobronchialiswhich has mantle and underpartsnot-yellow, wing-barsyellowandno black patchesis y1Y2b. This representation, it should be clearly understood, is tentative and approximate only. The characters are not really sharp, for there is much grading; but allowing for the effects of heterozygosis and for some actual breaking-up of factors I believe it gives a fairly correct view of the case. In particular we can see how it meets the difficulty which Chapman felt in acceptingleucobronchialisas in any sense derived frompinuswhich has a yellow breast, andchrysopterawhich has a black throat, seeing thatleucobronchialishas neither. We now recognize at once that this form could be produced by ordinary re-combination of the absence of Y1with the absence of B.
I note also with great interest that the modern observers agree that the so-called hybrids may have the song either of the one species, or of the other, or a song intermediate between the two. It may also be added that these two types have several times been seen, in the breeding season, paired with each other or with one of the other combinations.
Warblers
Allen[11]has described another excellent American example, the Tits of the groupBaeolophus bicolor-atricristatus. The formbicolorbelongs to the eastern States and ranges from the Atlantic coast to the Great Plains, andatricristatus, of east Mexico, extends from Vera Cruz to central Texas. In southern and central Texas the breeding ranges adjoin, and in this country various intermediates occur. The chief types differ in two main points.
The intergrades between the two have, as usual, received specific names. A detailed description is given by Allen, from which it appears that the gradation is very complete. In one case a series of 16 adults were all intermediates. It is not stated whether the collector took these at random, but from the local lists it is clear that the types are found not far away from the place where the intergrades were shot.
Another very striking case is that of the Tanagers, of the genusRhamphocoelus. In this group there are several local forms which are related to each other in remarkable ways. The forms known aspasseriniiandicteronotusexhibit the clearest phenomena of intergradation. The speciespasseriniihas a brilliant scarlet and black male, and it inhabits Honduras and Nicaragua. Proceeding southwards along the isthmus we find nextcostaricensiswhich has a male like that ofpasserinii(but a female with more orange than the olive-grey female ofpasserinii). Next we come to Panama which is occupied byicteronotus, sharply distinguished frompasseriniiby the fact that thescarlet is replaced by lemon-yellow. This sameicteronotusoccurs again as a pure type in Ecuador and many other parts of South America; but Colombia,between Panama and Ecuador, contains scarlets likepasserinii, yellows likeicteronotus, and various intergradesof several shades of orange. Thepasseriniimales from Nicaragua are indistinguishable from those of Colombia, and theicteronotusof Ecuador are the same as those in Panama. The orange intergrades, doubtless heterozygous forms, though collected at the same locality (Medellin in Colombia) as several pure yellows and pure scarlets, are in the British Museum series sorted out as a separate species under the namechrysonotus! Complications are introduced by the relations of these forms to another named type,flammigerus, but we may for our purpose leave that out of consideration, and say that the order of geographical sequence from Honduras to Ecuador is (1) scarlet, (2) yellow, (3) mixture of types, scarlet, yellow, orange, (4)yellow.
Similar examples exist in the birds of the old world, but I do not know of any that have been studied so fully as those of America. The best known is that of the two Rollers,Coracias indicuswhich spreads from Asia Minor through Persia, Baluchistan, the Indian Peninsula and Ceylon, andaffiniswhich ranges from Nepal, through Assam, Tenasserim and the Indo-Chinese countries. The two types are very different and may be distinguished as follows:
The wings are the same in both. In the provinces of Nepal, Sikhim, and Darjiling the two species coexist, with the result that intergrades have been frequently recorded. The line of intergradation extends to the coast, and birds showing various combinations of the two types from the Calcutta district exist in collections.[12]The case is interesting inasmuch as like that ofQuiscalusit shows a series of combinations of various metallic colours. Some of these are probably evoked by the development of pigment behind striations or other interferences already existing, but in the present state of knowledge it would be quite impossibleto suggest what the actual factors producing these appearances may be.
There are, naturally, many other cases among birds which are suspected of being in reality comparable, but in most of them the evidence is still inadequate. Among Lepidoptera also there are a few of these; perhaps the most striking is that ofBasilarchia "proserpina."[13]The genus is well known to European collectors under the nameLimenitis, of which we in England have one species,L. sibylla, the "White Admiral." A species very likesibyllain general appearance is common in the northern parts of the United States, ranging through Canada and Northern New England, but rarely south of Boston. This species has the conspicuous white bands across both wings like oursibylla.
There is also a more Southern type known asastyanax, which is very different in its appearance, being without the white bands and having a broad irroration of blue scales on the posterior border of the hind wings. The two are so distinct that one would not be tempted to suspect any very close relation between them. In its distributionastyanaxis described by Field as replacing arthemis south of latitude 42°. About Boston it is much more common thanarthemis.
The two forms encroach but little on each other's territory, but where they do coexist, a third form, known asproserpina, is found which is almost intermediate, with the white bands much reduced. There is now no doubt that thisproserpinais a heterozygous form, resulting from a combination of the characters ofarthemisandastyanax. Field succeeded in rearing a brood of 16 from aproserpinamother caught wild which laid 31 eggs, and of these, nine (five males, four females) resembled the mother, beingproserpina, and seven (four males, three females) werearthemis. There can be no question therefore that the mother had been fertilised by a malearthemisand thatno-white-bandis a factor partially dominant over thewhite band. Another point of interest which Field observed was that theproserpinafemale refused to lay on birch, poplar or willow, but acceptedwild cherry (Prunus serotina) a species on whichastyanaxcan live, though that tree is not known to be eaten byarthemis. Incidentally also the observations show that sterility cannot be supposed to be the bar which maintains the distinctness ofarthemisandastyanax.
In this connectionPapilio oregoniaandbairdiishould be mentioned.[14]P. oregoniais one of the numerous forms likemachaon, but rather paler. It is a northern insect, inhabiting British Colombia east of the Cascade Range, and reaching to Colorado.P. bairdiiis a much darker butterfly, representing theasteriasgroup of the genusPapilio. Likeasteriasit has the abdomen spotted at the sides, not banded as in themachaongroup. It belongs to Arizona and Utah extending into Colorado. From Colorado the formbruceiis described, more or less intermediate, likebairdiibut with the abdomen banded as inoregonia. W. H. Edwards records the results of rearing the offspring of thebairdii-like and of theoregonia-like mothers. Each was found able to have offspring of both kinds, that is to say,bairdiifemales gave both forms, andoregoniafemales gave both forms. It is not possible to say which is dominant, since the fathers were unknown. On general grounds one may expect that thebairdiiform will be found to dominate, but this is quite doubtful.
From this particular discussion I omit reference to those examples in which the permanently established types are obviously associated with special conditions of life. Where considerable climatic differences exist between localities, or when we pass from South to North, or from the plains into Alpine levels we often find that in correspondence with the change of climate there is a change in the characteristics of a species common to both. When I say "species" in such a connection I am obviously using the term in the inclusive sense. Some would prefer to say that in the two sets of conditions tworepresentative speciesexist. Whichever expression be preferred it is plain that such examples present another phase of the problem we have been just considering, and in them also we have an opportunity ofobserving the consequences of the overlap of two closely related types, but there are advantages in considering them separately. In the examples hitherto given, with the possible exception of the Papilios,[15]the two fixed types severally range over so extensive a region that it may fairly be supposed that in the different parts they are subject to considerable diversities of climate. There is no outstanding difference that we know distinguishing the habitats of the two forms; but in comparing Alpine with Lowland forms, or essentially northern with essentially southern forms we do know an external circumstance, temperature, that may reasonably be supposed to have an influence, direct or indirect, on the population.
Climatic Varieties
In this chapter we will examine certain cases which illustrate phenomena comparable with those just considered, though as I have already indicated, they form to some extent a special group. The outstanding fact that emerges prominently from the study of the local forms is that when two definite types, nearly allied, and capable of interbreeding with production of fertile offspring, meet together in the region where their distributions overlap, though intergrades are habitually found, there is no normally or uniformly intermediate population occupying the area of intergradation. Such phenomena as these must, I think, be admitted to have great weight in any attempt to construct a theory of evolution. True we must hesitate in asserting their positive significance, but I see no escape from the conclusion that they throw grave doubt on conventional views. Again and again the same question presents itself. IfAandBlately emerged from a common form why is that common form so utterly lost that it does not even maintain itself in the region of overlapping? Almost equally difficult is it, in the cases which I have numerated, to apply concrete suggestions based on any factorial scheme. We may see that inHeliconius eratothe type with the red mark on the hind wing probably contains a dominant factor, and that where the red mark is absent the metallic colours are exposed; and that similarly the green metallic colour may have another factor which distinguishes it from the blue. In this way we can fairly easily represent the various types oferatoon a factorial system as the result of the various possible combinations of two pairs of factors. But there we stop, and we are quite unable to suggest any reason why one area should have the red and the green type while another should have the blue also. So again withColaptesor the Warblers. By application of afactorial system, admittedly in a somewhat lax fashion, the genetic interrelations of the types can be represented; but how it comes about that each type maintains a high degree of integrity in its own region we can only imagine. Each has in actual fact a stability which the intermediate forms have not, but we cannot yet analyse the nature of that stability. Mendelian conceptions show us how by segregation the integrity of the factors can be in some degree maintained, but not why certain combinations of factors should be exceptionally stable. All that is left us to fall back on is the old unsatisfying suggestions that some combinationsmayhave greater viability than others, that theremaybe a tendency for like to mate with like, and so forth.
These difficulties acquire more than ordinary force in those cases in which the two fixed types inhabit regions differing in some respect so obvious and definite that we are compelled to regard each type as climatic and as specially adapted to the conditions. When for example an animal has a distinct type never met with except in Arctic or Alpine conditions, and another type proper to the plains and temperate regions, what are the characteristics of the population of intermediate latitudes or at intermediate levels? Some of the examples discussed in the last chapter may be instances of this very nature, but even if they are not, others are forthcoming which certainly are. The evidence of these cases leads to the suspicion that with further knowledge they will be found to consist of two classes, some in which the observer as he passes from the one climate to the other will find the intermediate area actually occupied by a population of intermediate character, and others in which, though we may presume the maintenance of intermediate conditions in the transitional area, there is no definite transitional population. This interrupted or discontinuous distribution seems, so far as I have means of judging, to be by far the more common of the two. I do not doubt that by sufficient search individuals representing every or almost every transitional form can be found, but it is apparently rare thatpopulationscorresponding to these several grades can be seen. The question has in few if any cases been studied with precision sufficient to provide a positive answer;but I suspect that real and complete continuity, in the sense thus defined, will only be found where the character of the local populations dependsdirectlyon the conditions of life, and shows an immediate response to changes in them apart from that postponed response which we suppose to be achieved by selection. Obviously the character must be one, like size for instance, capable of sensibly complete gradation.
The only example I have met with of the phenomenon of anything like a complete intergradation between local types really distinct in kind is that provided by the butterflyPararge egeria. It is well known to entomologists that this insect exists in two very different types, a northern one, the "Speckled Wood" of England, in which the spots are a pale whitish yellow, and a southern type having the full fulvous colour that we know as characteristic ofmegaera, the "Gatekeeper." It appears that Linnaeus gave the nameegeriato the southern type,[1]and our own is now calledegerides. Broadly speaking, so far as Great Britain, France, and the Spanish Peninsula are concerned, the tawny-colouredegeriaoccupies Spain and western France up to the latitude of Poitiers and the pale yellowegeridesextends from Scotland, where it has a scanty distribution, through southern England, where in suitable localities it is common, and the north of France to Paris.[2]The two types when placed side by side are strikingly different from each other, and are an excellent illustration of what is meant by climatic variation. The insect is not a great traveller and probably scarcely ever wanders far from its home. It should therefore be possible by collecting from north to south to find out how the transition is effected, whether suddenly or gradually. This at various times I have endeavoured to do, but I am still without exact information as to the population in certain critical areas. In addition to the information derived from specimens which I have collected or seen in the collections of others there is a good account of the general distribution in Europe given by the Speyers,[3]who evidently paid more attention to the subject than most lepidopterists have done, and many more recent records. In particular Oberthür[4]has published many details as to the distribution in western France and I am especially indebted to Mr. H. Rowland-Brown for a long series of notes as to the distribution in France generally, and to Mr. H. E. Page and Dr. T. A. Chapman, Mr. Oberthür Prof. Arrigoni degli Oddi, Mr. H. Williams and other correspondents, for showing me forms from many localities. The butterfly is attached for the most part to woods of deciduous trees and to country abounding in tall hedges or rough scrub. It is not usually to be found in highly cultivated districts or in very dry regions. Hence there is necessarily some want of continuity in the distribution at the present time and I should think a mile or two of arable land without big hedges would constitute a barrier hardly ever passed. The larva feeds on several coarse grasses, especiallyDactylis glomerata. Barrett mentions alsoTriticum repens. In this country the winter is usually passed in the larval stage, but I have found that in captivity, at least, there is much irregularity. The larvæ feed whenever the weather is not very cold and may pupate, but if sharp cold comes on when they are pupating or nearly full-grown they often get killed unless protected.
Some writers speak of a difference between the early and later broods, but I have never noticed this, and I do not think that the general tone of the yellow is affected by the seasons (see Tutt,Ent. Rec., IX, 1897, p. 37).[5]
Beginning at the south of Spain the thoroughly fulvous typeegeriais common at Gibraltar in the Cork woods, at Granada, and doubtless generally. Lederer is said to have found only this type in Spain (Speyer), and though I have no precise information as to other places in the Peninsula north of Jaen I feel tolerably sure that there is no change from south to north.[6]Immediately north of the Pyrenees we still meetegeriaexclusively, and up to Poitiers at least there is no noticeable change. But somewhere between Poitiers and the bottom of the Loire valley at Tours, the genuine southern type comes to an end, and the whole population begins at the Loire to be of an intermediate type, easy to distinguish both fromegeriaand fromegerides. As to the exact condition of the species in the fifty miles separating St. Savin on the Vienne from places on the Loire I have no adequate information. I have only one small sample from there, but it does contain insects both of the southern and intermediate types taken on the same day, in a wood near Preuilly. Oberthür also states that at Nantes the true southern form exists in company with the northern. From this I infer that the southern form extends up the coast further than it does inland, but I imagine the representative spoken of as northern would be of usual Brittany or intermediate type.
The Vienne river joins the Loire, so the true southern type reaches over into the basin of the Loire. From the Loire (Tours, Corméry) north to Calvados (Balleroy) only the intermediate is found, so far as I know, and the same type extends over Brittany.[7]In general, however, the woods near Paris have the thoroughly northern typeegerides, but at St. Germain-en-Laye and at Etampes (Oberthür) the population approaches the intermediate type.
On the whole the intermediate type is certainly less homogeneous than either of the extremes, and females with the two central spots either paler or more fulvous than the rest are not uncommon, but I have never taken one on the Loire or in Brittany which I should class with either of the extreme types.
Before speaking of the distribution in other parts of France and in Europe generally I will briefly state the results of my breeding experiments. The work was done many years ago before we had the Mendelian clue, and it is greatly to be hoped that some one will find opportunities of repeating it. Crossing the English and the thoroughly southern type the families producedagree entirely with the intermediates of Brittany and the Loire. Reciprocals are alike. Of F2I only succeeded in raising very few and of those that I had (about 30) nearly all were intermediate in character, though perhaps rather less uniform than F1. One family alone, containing only 4 specimens, had oneegerides, and three fulvous intermediates. As the case stands alone I hesitate whether or not to suppose it due to some mistake. Moreover from F1crossed back with the respective parental types I had fairly long series, especially from F1× the southern type, and looking at these families I cannot see any clear evidence of segregation. On the contrary, I think that though there are slight irregularities, they would, taken as a whole, be classed as coming between the intermediate type and the extreme form used as the second parent. This at least is true when the second parent was of the southern type.
On this evidence I have regarded the case as one in which there is no good evidence of segregation and as conforming most nearly with the conventional view of gradual transition in response to climatic influences. Such influence must however be indirect; for I reared five generations of the northern type in England, and these, though they included several abnormal-looking specimens in the last generation and then died out, did not show any noticeable change from the fulvous colour of the wild type. Merrifield[8]also found that heat applied to pupae of the northern type produced no approach to the southern type.
Looking at the facts now in the light of more experience it seems to me just possible that the case may be one in which, as in Nilson-Ehle's Wheats, the dominant differs from the recessive in having two pairs of factors with similar effects. The fulvous type for example may have two or more elements in separate pairs which together produce the full effect, and the intermediate may have one of these. If this were so, some segregation should of course eventually be observable, but the proportion of the various fulvous and fulvous-intermediate individuals would be large, and the reappearance of actualrepresentatives of the northern type might be rare. I admit that this is a somewhat strained interpretation of the facts, and as yet it is not entitled to serious consideration. Nevertheless I am led to form some such expectation partly from the great difficulty in the way of any other, partly from the evidence of the small mixed sample found at Preuilly and partly from the statements given by Oberthür. There are moreover other features in the general distribution of the species which make it improbable that the dependence on climate can after all be so close. Published lists are unfortunately of little use in deciding which form occurs at a particular place, because, since the nameMeonehas ceased to be used for the southern form, there is no complete unanimity among authors as to the application of the namesegeriaandegerides, and unless more particulars are given, either name may be used for either form. Besides this, difficulty arises from the fact that the intermediate type is not generally distinguished at all, and English collectors finding it, may easily record it as the southern type. From Staudinger's note on the distribution, I gather that he, on the contrary, reckoned the intermediate with the northern type, as do the Speyers also. The late Mr. J. W. Tutt was careful to distinguish the three forms and has left several useful records. Easy therefore as it might seem to be to make out the distribution of such a familiar insect in its various modifications, there are serious practical difficulties, and until long series are brought together with this special object in view many obscurities will remain.
With only the series from England, the west of France, and Spain before one it would be easy to regard the successive series of tones as a fair measure of climate; the brighter the colour, the hotter might one expect the locality to be. Such rough correspondence is often to be observed in butterflies and birds. It becomes impossible to take these simple views in the light of more complete knowledge. Beginning with France the fulvousegeriaoccupies the lower valley of the Rhone, probably from well above Lyon, though I have no exact information respecting the country above Avignon. According to Speyer it also takes the department of Lozère. The same authority says that Puy-de-Dômehas "egeria," meaning perhaps the intermediate form, with the fulvous form much less commonly. Next comes the curious fact that though the Lower Rhone (Avignon, Tarascon, Nîmes) has the true fulvous form, Hyères, Cannes, Grasse, Nice, Digne, and Alassio havethe intermediate. Savoy has the intermediate (Chambéry) and evenegeridesperhaps, though in the same latitude on the west of France there is nothing but the fulvous type. At Chalseul and Besançon (Doubs) the ordinary northern type is found. Switzerland generally, I believe, has the northern type, but Staudinger givesegeriafor Valais and the intermediate occurs in Vaud.[9]The south side of the Alps has probably colonies of the paleegerides, and of intermediates. Orta, with a very hot summer, has the English type (Tutt,Ent. Rec., XII, 1900, p. 328). Locarno has the intermediate (ibid., XV, 1903, p. 321). North Italy in general and western Piedmont have the intermediate; but further southegeriabegins, at what region I do not know. Speyer gives on his own authority the remarkable statement that at Florence both extremes occur, but chiefly intermediates between the two. Mr. R. Verity however kindly informs me that in his experience this is not so, and that neither the real southern type nor the northern occur there. Sardinia, Sicily, Crete all have the southern type. Greece probably has various types. Staudinger (Hor. Ross., VII, 1870, p. 78) says intermediates resembling Nice types common everywhere, but from "Greece" the British Museum has a series that would pass for English specimens; and the same type occurs near Constantinople. The island of Corfu has a pale intermediate, distinct fromegeridesbut approaching it. In Roumania all three forms are recorded from various places:egeriain the Dobrutscha; not quite typical (presumably an intermediate) at Bukharest; intermediate in various mountainous localities as well as in Macedonia and Dalmatia; butegeridesin Azuga at about 3,000 feet.[10]Hungary has the trueegeridesalso. (Cf. Caradja,Deut. Ent. Zt., IX, p. 58.) Mathew records the samefrom Gallipoli (E. M. M., 1881, p. 95). Staudinger does not distinguish the intermediates from the northern, but he gives "egerides" for Armenia and Fergana (Central Asia). As against the mere proximity of a great mountain chain being the influence which keeps the Riviera population intermediate may be mentioned the fact that the northern foothills of the Pyrenees have the pure southern type, and the climate of Cambo must surely be far cooler than that of Nice. The exact locality of the Greek specimens is not given, but there can be no part of Greece which is not much hotter in summer than Brittany, or Calvados, which have the intermediate, not the English type.
In face of these facts it can scarcely be maintained that average temperature is the efficient cause of the particular tone of colour which the butterfly shows in a given region. Nevertheless it is clear that climate counts for much in determining the distribution. It is noticeable that though the paleegeridescan be established in a warm climate we never findegeriain cold climates, and even the intermediate is not found in places that have a hard winter. I suspect that the distribution of the broods through the year and the condition of the animal at the onset of hard frost are features which really determine whether a strain can live in a particular place or not. Though the truth of the suggestion cannot be tested by experiments in captivity, which at once introduce disturbances, I incline to the idea thategeriahas not got the right periodicity for northern climates. If it could arrange its life so that the population consisted either of young larvae, or perhaps of thoroughly formed pupae[11]at the onset of winter, it might, for any obvious reason to the contrary, be able to live in England. It is irregularly "polyvoltine," as the silk-worm breeders say, and as soon as a little warmth encourages it, a new generation starts into being, which if the frost comes at an untimely moment, is immediatelydestroyed. Many species are continually throwing off individuals which feed up fast[12]and emerge at once if the temperature permits, and I imagine a species of Satyrid wholly or largely represented by such individuals could scarcely survive in a country which had a hard winter. For such a climate some definite periodicity in the appearance of the broods may well be indispensable. But assuming thategeriais cut off from cold climates for such a reason, there is nothing yet to connect these habits with the fulvous colour, and until breeding can be carried out on a satisfactory scale there is no more to be said.
From time to time records appear of individual specimens more or less fulvous being caught in southern England, especially in the New Forest.[13]It would be interesting to know what offspring such individuals might produce. From the evidence now given some notion both of the strength and the weakness of the case considered as one of continuous climatic variation can be formed. I know no other equally satisfactory. Whether or not definite mixture of the intermediates with either of the extremes will be proved to occur, the case differs materially from those considered in the last chapter in the fact that at all events there is no general overlapping of forms. In a species so little given to wandering, overlapping could indeed scarcely be expected to occur. It is this circumstance which makes the species preeminently suitable as a subject for the study of climatic influences, and I trust that entomologists with the right opportunities may be disposed to explore the facts further.
Just as many species, likeegeria, have varieties which can be regarded as adapted to northern and southern regions, so there are also several which have lowland and Alpine forms quite distinct from each other. Every such case presents an example of the problem we have been considering. As the collector passes from the plains to the Alpine region, how will he find thetransition from one form to the other effected? Does the lowland form give place to the Alpine form suddenly, with a region in which the two are mixed, or will he find a zone inhabited by an intermediate population? I have spent a good deal of time examining the facts in the case ofPieris napiand its Alpine female varietybryoniae, and though there are many complications which still have to be cleared up, no doubt is possible as to the main lines of the answer. If in any valley in the Alps inhabited by bothnapiandbryoniaethe collector catches every specimen he can, beginning at the bottom and working up to 7,000 feet, he will at first get nothing butnapi. At about 2,500 feet, he may catch an occasionalbryoniaeflying with thenapi. After 3,000 feetnapiusually ceases, and onlybryoniaeare found. As an exception a colony ofnapimay be met with at much greater heights. I once found them in numbers at about 6,000 feet.[14]Not only were they free from any trace of modification in the direction ofbryoniae, but they were of the thoroughly southern type ofnapi, being a late brood of that large and very pale kind (meridionalis) almost destitute both of dark veining above and of green veining below, which are common on the shores of Lago Maggiore and in other hot southern localities. Not far off at the same level were typicalbryoniaein fair abundance. Occasionally an intermediate may be met with. I have taken a few, for example, at Macugnaga and at Fobello. These, however, in my experience are rarities in the Alps. Fleck[15]gives notes on the distribution in Roumania which shows the same state of things. The lowland form is not transformed though found at great heights, and at Azuga (nearly 3,000 feet)bryoniaeoccurs with only occasional "flavescens," viz., intermediates of the second brood.
If this were all the evidence we should be satisfied that the lowland and Alpine types keep practically distinct, overlapping occasionally, but rarely interbreeding. The problem would remain, how is the distinctness of the two types maintained in the region of overlapping? Nowadays, I suppose, we shouldincline to answer this question by reference to segregation, and perhaps by an appeal to selective mating. The suggestion that segregation does take place is certainly true to some extent. There are, however, difficulties in the way, and the whole subject is one of great complexity. My own experiments were made in pre-Mendelian times and were not arranged with the simplicity which we now know to be essential. The results are neither extensive enough nor clear enough to settle the many collateral questions which have to be considered, and the work ought to be done again. Nevertheless, some notes of the observations may have a suggestive value.
When I began, I did not sufficiently appreciate that the "napi" group, omitting the North American forms, and the Asiatic representatives, has at least three chief types in western Europe. The differences we have to deal with are manifested by the females only, so in this account particulars as to the males are omitted for the most part. These are (1) our own Britishnapi; (2) the form found in the south, from the Loire downwards, and in the Italian Alps, which I think may be spoken of asmeridionalis; (3)bryoniae, which is a form clearly recognizable in thefemaleonly, and is found only in the arctic regions and in the Alps above 2,500 feet. The first two have several broods, two, three, or more, according to opportunity, and the first brood is different from the later ones. Innapithe markings on the upper surface are a dark grey but inmeridionalisthey are a pale silvery grey and much less extensive. In the later broods ofnapithere is much less general irroration of the veins, and the spots stand out as more defined and blacker. These differences vary greatly in degree of emphasis. Inmeridionalisthe later broods are entirely different from the first. Instead of having silvery markings they have the ground colour quite white, with the spots large and a full black. On the under side of the hind wings the usual green veins are almost absent, and I have seen individuals which could scarcely be distinguished fromrapae. To these later broods the termnapaeaeis sometimes applied, but I here usemeridionalisfor the southern race in general as applicable to all broods.