494 



NATURE 



[September 17, 1908 



ancestrulae of a given species still carry the determinants 

 representing those allelomorphs. In species in which the 

 vicarious avicularia are of normal occurrence there is no 

 difficulty in this hypothesis. In others, of which examples 

 may be found in Schizoporella, the vicarious avicularia 

 make their appearance rarely, in a very small proportion 

 of colonies. In these cases 'the facts might be accounted 

 for on the hypothesis of the chance recombination of 

 allelomorphs which are ordinarily separated, unless, indeed, 

 it should prove to be the case that the vicarious avicularia 

 represent a recessive character which is usually prevented 

 from making its appearance by some dominant factor. 



A single series of cases of this kind will not carry con- 

 viction, but there are many facts with regard to the 

 distribution of adventitious avicularia that may point in 

 the same direction. We may recur to the fact that the 

 form of these appendages may be eminently characteristic 

 of a whole series of species which from their similarity 

 in other respects are naturally associated in a single genus 

 or family. The most striking instance of this is, perhaps, 

 the genus Bugula, in which we find the avicularium par 

 excellence. The variations of this type of avicularium are 

 comparatively slight, and for the 'most part depend on 

 differences in position with regard to the zocecia and on 

 minor modifications of size, shape, and length of stalk. 

 Both in Bugula and in the allied genus Bicellaria the 

 avicularian characters may be described as relatively con- 

 stant ; and since they belong to a type that is rarely met 

 with in other genera, they seem to' confirm the evidence 

 afforded by other structural features that the species which 

 possess them are related to one another. But even in 

 Bugula, where the avicularia reach the summit of their 

 development, we meet with species or varieties in which 

 these appendages are invariably absent throughout the 

 colony. This may be illustrated bv Bugula neritina, a 

 widely distributed species which in the Mediterranean and 

 certain other districts is remarkable for the complete 

 absence of avicularia, although in other structural features 

 it shows a close affinity to other species of Bugula. In 

 Australian and Oriental waters, however, there occur 

 forms which can hardlv be distinguished from B. neritina 

 except by the fact that they always possess numerous 

 avicularia of the specialised character that is so distinctive 

 of the genus. It does not matter for our present purpose 

 whether these are to be regarded as a varietv of B. neritina 

 or not. If the appearance of avicularia mav be regarded, 

 on Mendelian principles, as due to the presence of one or 

 more allelomorphs, it is possible to understand that these 

 may be omitted in certain cases, and that there may thus 

 be a close affinity between two forms, one of which differs 

 from the other in what appears at first sight so essential 

 a respect «s the complete absence of the avicularia, which 

 W'e are justified in regarding as the most important feature 

 of the genus. 



A second case of the same general nature mav also be 

 noticed. In the family Cellulariida; are included 'a number 

 of delicate erect species which are commonly placed in 

 the genera Caberea, Scrupocellaria, Menipca.'and Cellu- 

 laria. _ The first two of these are distinguished by possess- 

 ing vibrncula as well as avicularia. Menipea is defined 

 as possessing avicularia, but no vibracula ; while Cellularia 

 peachti does not possess either kind of appendage. A 

 species known as Amastigia nuda has been placed in a 

 separate genus because of the absence of vibracula and 

 their replacement by avicularia, w^hile in other respects it 

 agrees with Caberea, in which the vibracula reach a 

 development not exceeded by those of any other Cheilo- 

 stome. Before considering the bearing of' these facts we 

 nriay appropriately consider another instance taken from 

 the same family, although bv doing so we are for the 

 moment leaving the question of the avicularia. In the 

 genera Caberea, Scrupocellaria, and Menipea certain 

 species are distinguished bv having the free surface of the 

 zooecium protected by a peculiar spine known as the 



scutum," which is usually flattened and much expanded 

 at Its free end, where it overarches the membranous frontal 

 surface in such _ a way as to cover and presumably to 

 protect It. But in each genus other species are char.acter- 

 ised by the complete absence of the scutum, while in others 

 " w'^"'"'' '" ^•'^'■>''"S degrees of reduction, 



We have thus several cases in which certain species 



NO. 2029, VOL. 78] 



differ from their near allies in the complete absence of a 

 structure which is, as a rule, one of the most distinctive 

 features of the genera to which they are respectively 

 assigned. Should it be possible to prove that the appear- 

 ance of the organ in question, whether avicularium, 

 vibraculum, or scutum,' was of the nature of an allelo- 

 morphic character, its disappearance would be readily 

 intelligible. 



The facts which I have indicated with regard to the 

 so-called Cellulariidae have not hitherto been sufficiently 

 discussed ; but I imagine that most systematists who have 

 considered the question have assumed that the scutum, 

 for instance, has undergone parallel evolution in Caberea, 

 Scrupocellaria, and Menipea, either having been independ- 

 ently evolved in each of the three cases (a most improbable 

 supposition), or having independently undergone a series 

 of regressive changes of precisely similar character in the 

 three genera. 



But it is perhaps in the mode of occurrence of adven- 

 titious avicularia that we find the strongest reason for 

 believing in the existence of some form of alternative j 

 inheritance. We may indeed go so far as to assert that 

 alternative development does actually take place, whether 

 the explanation of the facts is given by the Mendelian 

 theory or not. The difference between the pointed and the 

 round avicularia is a very definite one, which — it is no 

 exaggeration to say — may be observed in hundreds of 

 species. When these species are arranged under genera 

 according to the result of a study of the whole of the 

 evidence derived from all the characters that have proved 

 valuable in classification, w-e find that many genera include 

 some species with one type of avicularium and others with 

 the other type. It should perhaps be pointed out that the 

 validity of many of these genera is a matter on which 

 differences of opinion exist. The subject is undoubtedly 

 a difficult one, and we are far from having arrived at any 

 certainty with regard to the classification of the Cheilo- 

 stomata. But it is perfectly certain that we could not 

 utilise the two kinds of avicularia in dividing these 

 Polvzoa into two main series, since there are innumerable 

 cases in which both kinds occur in a single colony. This 

 is a fact to which I shall return later. 



We may accordingly maintain that, although much is 

 probably faulty in our present system, we have clear 

 evidence that the same genus may include species which 

 differ in the type of avicularium ; and, moreover, that these 

 are not exceptional, but, on the contrary, are of common 

 occurrence. .'\ few instances will make these points clear. 



In the encrusting species and in certain others the 

 avicularia commonly occur, as we have already seen, in 

 a position near the orifice of the zoa'cium, where they are 

 usually cither lateral or suboral. In one of the species 

 with lateral avicularia these appendages may be of the 

 pointed type, while in another they may be rounded ; and 

 the same statement may be made \vith regard to the sub- 

 oral avicularia. Within the limits of the same genus we 

 may further notice that certain species have lateral avicu- 

 laria, while others have suboral avicularia. Here, again, 

 we find the same indifference as to the shape of the rostrum 

 and mandible. 



If we mij,'ht provisionally suppose that the two kinds of 

 avicularia constituted an allelomorphic pair, represented by 

 .\a. and that the lateral and suboral positions indicated 

 a second allelomorphic pair, Bb, the four combinations, 

 .\B, Ab, aB, ab, would be theoretically possible. We 

 might, in other words, have pointed or rounded lateral 

 avicularia. and pointed or rounded suboral avicularia. All 

 these conditions actually occur in such genera as Lopralia 

 and Schizoporella ; and in some cases two species which 

 agree in the form of the avicularia but differ in their posi- 

 tion, or agree in the position but differ in their form, 

 appear on other grounds to be nearly related one to the 

 other. 



Other cases may be taken from Retepora, an instance 

 where we may feel ourselves on comparatively secure 

 ground, since there are strong reasons for believing the 

 genus to be a natural one. The genus as a whole t>ossesses 

 an almost bewildering variety in the form, position, and 



1 The ca-Je of the scutum is le=s sfrik'ns ihnn that nf the other structures 

 under consideration, since condit-ons interiuediate between full develop- 

 ment and complete absence are not uncommon. 



