September 17, 1908J 



NA rURE 



493 



in many cases to give but slight indications of affinities. 

 It cannot, for instance, be assumed, without furtlier 

 evidence, that two species possessing an identical type of 

 avicularium are nearly related. The complete absence of 

 avicularia in a particular species is no sufficient reason 

 for removing- that species from an assemblage of forms 

 in which avicularia are always present. And, lastly, there 

 may be good grounds for believing that two forms with 

 entirely different types of avicularia are closely related, and 

 in some cases may even belong to the same species. 



The result of a comparative study of the Cheilostomata 

 leads, in fact, to the conclusion that although certain 

 genera or species are characterised by the possession of 

 one or more definite types of avicularium or vibraculum, 

 other genera or species show no such constancy in this 

 respect. The occurrence of the same type of avicularian 

 appendage in the species of widely separated genera and 

 the diversity of type of avicularium within the limits of 

 a single genus or species render it most difficult to frame 

 any theory that will account for the facts. Are we to 

 assume that a given type of avicularium has been evolved 

 independently in a number of cases, or must we suppose 

 that species svith that type have inherited it from a common 

 ancestor? If the latter hypothesis be the correct one, we 

 seem to be led to the conclusion that the ancestral Cheilo- 

 stomes were provided with most of the types of avicularia 

 that actually occur in existing species, many of which 

 have lost one or more of those types. 



In trying to arrive at some conclusion with regard to 

 these points we may notice, in the first instance, one fact 

 which stands out with great distinctness — namely, that, 

 whatever the modifications of the avicularium may be, 

 the mandible is usually either acutely pointed at its free 

 end or rounded and spatulate at its termination. The 

 difference may at first sight appear unimportant, but I am 

 inclined to believe that it is an indication which may lead 

 us to results of great significance. 



Though it may be going too far to assert that all 

 avicularia belong to one of these two types, there is 

 usually no difficulty in recognising either the pointed or 

 the rounded character in every avicularium present on a 

 colony. The distinction may be observed by inspecting 

 the form of the rostrum in a dry preparation of a part of 

 the zoarium, but it is seen with more certainty when the 

 mandibles have been isolated and are examined in Canada 

 balsam. So striking is the difference that the inquiry 

 naturally suggests itself whether there is any indication of 

 the evolutionary meaning of the two kinds of avicularium. 

 It appears to me probable that a condition which is 

 characteristic of the existing genus Steganoporella may 

 furnish the answer to this question. In this genus 

 avicularia are typically absent, but in each species the 

 zooecia are of two kinds, distinguished by differences in 

 the shape and structure of the, opercula and orifices. The 

 anatomy of the zooecia is known in but few cases, but in 

 those that have been observed both kinds of zooecia possess 

 polypides. In one division of Steganoporella the more 

 differentiated zooecia show some resemblances to the 

 pointed type of avicularium, while in a second division 

 they more nearly resemble rounded avicularia. I am 

 inclined to believe that these conditions correspond re- 

 spectively with the two kinds of differentiated avicularia 

 of other Cheilostomes. 



The avicularia most commonly met with occur as 

 appendages of the ordinary zooecia, which alone constitute 

 the main framework of the colony. But in addition to 

 these, the " adventitious " avicularia of Busk, we find, 

 although less commonly, another kind known as the 

 " vicarious " avicularium, from the fact that it occupies 

 the place of an ordinary zooecium, with which it agrees 

 more or less closely in point of sire. Its mandible is 

 usually of the rounded type, appropriately referred to as 

 "duckbill-like," and is readily seen to represent the 

 operculum of an ordinary zooecium. Compared with this 

 the mandible and the orifice which it closes are greatly 

 enlarg-ed. while the occlusor muscles have become corre- 

 spondingly increased in size. The polypide is generally 

 absent in the vicarious avicularium. 



Pointed avicularia of the vicarious type occur normally 

 in the species of Onychocella, which, alike by their struc- 

 ture and by their early paU-eontological appearance, may 



NO. 2020, VOT,. 78] 



be regarded as representing a primitive type of the Cheilo- 

 stomata. Vicarious avicularia with a rounded mandible 

 occur in certain species which I refer provisionally to 

 Siphonoporella, as well as in a small proportion of the 

 species of Membranipora and Flustra. All these may 

 fairly be regarded as belonging to a comparatively un- 

 differentiated type of Cheilostomata, and their vicarious 

 avicularia are usually the only ones present. It is thus 

 not improbable that the avicularium in these cases really 

 represents an early stage of evolution. But we must notice 

 that precisely similar rounded vicarious avicularia make 

 their appearance occasionally in species of a much more 

 differentiated type, as in the well-known Schizoporclla 

 Cecilii ' and in certain other species which may for the 

 present be referred to the same genus. In the majority 

 of the very numerous species of Schizoporella vicarious 

 avicularia are not known to occur, and it is thus impossible 

 to regard them as a typical attribute of the genus. 



The vicarious avicularia, which by their position and 

 general structure are so easily comparable with tlie ordinary 

 zocecia, are usually supposed to represent an initial stage 

 in the evolution of the avicularium. But if this view 

 be correct, how are we to account for the sporadic way 

 in which these structures occur in a series of genera such 

 as Membranipora, Flustra, Schizoporella, and Cellepora, 

 the last two of which, at any rate, are highly specialised 

 in other respects? What conclusion can we draw from 

 the association, in one and the same colony, of this type 

 of avicularium with adventitious avicularia of the niost 

 specialised description ? How can we explain the fact that 

 each kind of avicularium occurs in certain species, but 

 not in all the species, of many distinct and not specially 

 related genera? And, lastly, what is the significance of 

 the fact that certain species of a genus which is normally 

 provided with avicularia may be totally destitute of these 

 organs? These are some of the problems of which no 

 satisfactory solution has at present been given. On the 

 ordinary view of the way in which the species of a genus 

 are interrelated we should perhaps not expect to find that 

 two species which are closely similar in other respects may 

 be distinguished by possessing entirely different types of 

 avicularia. 



I am aware of the fact that it is perhaps premature 

 to indulge in speculations which are unsupported by 

 experimental evidence. But it appears to me worth while 

 to suggest that some of our difficulties might be removed 

 by appealing to the results obtained by workers on 

 Mendelian inheritance. An essential part of the theory 

 Jiere involved is that in the formation of the gametes of 

 an organism there is a segregation of certain paired or 

 " allelomorphic *' characters whereby some of the gametes 

 are endowed with qualities by virtue of which they trans- 

 mit one of the characters, while the rest of the gametes 

 become capable of transmitting the characters of the other 

 member of the allelomorphic pair. It has recently been 

 made probable by Prof. Bateson, whose views have been 

 confirmed by others, that the actual appearance of a par- 

 ticular character may be dependent on a coupling of two 

 allelomorphs belonging to distinct pairs. If only one of 

 them is present the character will not show itself. The 

 phenomenon of reversion on crossing is thus e.xplained as 

 due to the combination of allelomorphs present in the 

 isolated condition in two parental forms. 



Is it not possible that the perplexing occurrence of 

 vicarious avicularia in some of. but not by any means in 

 all, the colonies of certain species may be interpreted as 

 .1 reversion due to the combination of two or more allelo- 

 morphs that may not have occurred together in the 

 parental forms? We have seen that there is some reason 

 to believe that these avicularia are really of an archaic 

 character, from their occurrence in certain genera of a 

 primitive type, known in some cases by palaeontologicaf 

 evidence to have appeared early in the evolution of the 

 Cheilostomata. We may further remember that we hav^ 

 distinct evidence that Cheilostomes of a differentiated type 

 may retain certain primitive characters, in the occurrence 

 of a Membranipora-like form of ancestrula in so many of 

 them. If, then, we may suppose that the appearance of 

 vicarious avicularia is due to a combination of more than 

 one allelomorph we may recognise the possibility that the- 

 ' Kirkpatrick, " Ann. Mag. Nat. Hist." (6), v., 1890, p. 21. 



