September 17, 1908] 



NATURE 



495 



size of the avicularia, among which, however, we may 

 distinguish the following kinds : — (i) The suboral avicu- 

 larium, closely related to the orifice and usually termed 

 " labial," because it occurs on what may be described 

 as the lower lip ; (ii) frontal avicularia, on some part of 

 that surface of the zooecium which bears the orifice ; 

 (iii) basal or dorsal avicularia, on the backs of the 

 branches ; (iv) fenestral avicularia, which guard the edges 

 of the fenestrae or meshes of the colony. 



In many of the species of this large genus the suboral 

 avicularia are of the small rounded type. In other species 

 they are small and pointed, with an acute mandible ; while 

 others are distinguished by possessing suboral avicularia 

 that may be described as gigantic. 



Among the frontal avicularia similar differences exist. 

 In one case that has come under my observation a re- 

 markable variation of this kind is found within the limits 

 of a single species. Remembering the great difficulty 

 there often is in arriving at certainty with regard to the 

 limits of the species in the genus under consideration, I 

 wish to emphasise the fact that this instance is taken 

 from Retepora phoenicea, a form that not only has well- 

 marked specific characters of the ordinary kind, but is 

 remarkable in having a beautiful carmine-red or violet 

 colour, a respect in which it differs from most of its 

 nearest allies. The frontal avicularia of this species are 

 usually of the pointed type, but in the variety in question 

 — a colonv from Torres Straits — they are, so far as I 

 ha%'e been able to ascertain,' all of the rounded kind. 



The fenestral avicularia show a similar behaviour. In 

 South .Australian waters there are a number of forms 

 which are regarded as varieties of Retepora monilifera. 

 In the form known as var. miinita there is usually a 

 suprafcnestral avicularium of large size, distinguished by 

 having a rounded mandible, which is a good deal broader 

 than it is long." In another form of the same species, 

 distinguished by MacGillivray as var. acutirostris, the 

 i)in?ii7n-avicularium may either occur as such in some of 

 the fenestrae, or be replaced in others by a large avicu- 

 l.irium of the typical pointed form. 



In other species a gigantic infrafenestral avicularium 

 commonly occurs, but while these structures are found in 

 a considerable proportion of the fenestrje of some colonies 

 they appear to be completely absent in other colonies. In 

 this series of cases, which is well illustrated by Retepora 

 phoenicea, I think there is clear evidence that different 

 colonies, from the same locality and belonging to the same 

 species, may show the two conditions of presence and 

 absence respectively of fenestral avicularia. According to 

 the ordinary criteria by which species of Polyzoa are dis- 

 criminated, it might be necessary to place these in different 

 species — a result which is not supported by other evidence. 

 I think we must conclude that a species may have the 

 facultv of entirely dropping out some complete series of 

 organs, like certain kinds of avicularia. The Mendelian 

 principle may here come to our aid by showing the theo- 

 retical possibility of having the two conditions represented 

 in a series of colonies of identical parentage. If this 

 should really be the explanation of the facts, it should 

 occasion no surprise if some members of the immediate 

 progeny of a colony in which a certain type of avicularium 

 is absent should be found to be provided with a complete 

 armature of these appendages. 



The cases so far considered may conceivably be explained 

 oil ordinary Mendelian lines by assuming that an entire 

 colony is homozygous or heterozygous with regard to par- 

 ticular characters. Remembering that the so-called 

 ancestrula, or primary individual, does not show all the 

 characteristics of the mature colony, we must, however, 

 assume that the determinants present in it do not find their 

 full expression until the budding orocess has commenced. 



But we are by no means at the end of our difficulties, 

 even in considering the distribution of the appendages we 

 have so far discussed. The instances already given have 



1 It may tie noted that it i^ extremely difficult and often impossible to 

 make a study of every part of a lar^e .-ind irreeular Retepore sufficiently 

 exhaustive to iustify one in asserting positively that all pans are identical 

 in respect of their avicularia. 



- This characteristic 7HK«;7rt-avicularium is probably merely fln enlarged 

 f rm of the small circular type of avicularium m^t with as labial avicularia 

 and in other positions in many species. 



NO. 2029, VOL. 78] 



for the most part been cases in which an entire colony 

 differs in certain respects from other colonies. We have 

 still to notice the common case in which there are differ- 

 ences in different parts of one and the same colony. No 

 theory can be considered complete unless it is able to 

 account for these differences. 



I approach this part of the subject with great trepida- 

 tion, conscious as I am of the absence of experimental 

 evidence for the suggestion I wish to make. This sugges- 

 tion is, briefly, that if a segregation of characters norinally 

 takes place in the formation of the gametes of an organism, 

 it is conceivable that an analogous segregation may occur 

 in the blastogenic processes, or, in other words, in the 

 formation of a bud. It may be asserted positively that 

 there is a very definite differentiation of individuals at 

 this time, not only in the Polyzoa, but also in other animals 

 which increase by budding. The fact that some of these 

 differentiations appear to be alternative suggests the possi- 

 bility that they are due to a process which resembles the 

 Mendelian segregation of determinants in the gametes. 



One of the instances which appears to me specially 

 suggestive in this connection is the genus Steganoporella, 

 the species of which are remarkable for the dimorphism of 

 their zocecia. This dimorphism is expressed, as we have 

 already seen, by differences in the opercula and in their 

 muscles, and in' the form of the orifices which are closed 

 by the opercula. It is not too much to say that every 

 individual in a Steganoporella colony belongs to one of 

 the two types in question ; and, so far as I am aware, 

 intermediate forms of zocecium do not occur. It is thus 

 a positive fact that the blastogenic tissues undergo some 

 sort of differentiation of an alternative character, and there 

 is at present no reason for believing that the differentiation 

 is in any way correlated with the production of se.xual 

 cells by either of the two kinds of zooecia. 



Another case which seems to me specially suggestive is 

 that of the simultaneous occurrence in the same colony 

 of two different kinds of avicularia. These instances^ are 

 not confined to a few species, but may be found in a 

 number of genera which do not constitute a single 

 assemblage of related forms. The pointed and rounded 

 adventitious avicularia may be scattered about promiscu- 

 ously in the same colony, or even on the same zocecium. 

 Sometimes avicularia of one of the two types norinally 

 occur in a particular position, but are occasionally replaced 

 by avicularia of the other kind, an example of a general 

 phenomenon to which Prof. Bateson has given the name 

 of " homceosis." 



Excellent illustrations of this substitution may be taken 

 from the genus Retepora. In the R. monilifera series 

 already considered, the munita and acutirostris types of 

 avicularia mav occur in different fenestr.-e of the same 

 colonv. R. gr'anulata usually possesses a labial avicularium 

 and a frontal avicularium, both of the small rounded kind. 

 In one of the colonies of this species dredged by the 

 Siboga most of the labial avicularia are of this type, but 

 a certain proportion of the zocecia have a pointed labial 

 avicularium. In another colony most of the frontal avicu- 

 laria are small and round, but in some of the zocecia they 

 are large and pointed. In both instances the examination 

 of the mandibles proved the reality of the distinction 

 inferred from the shape of the calcareous parts. 



Instances of a similar substitution could easily be multi- 

 plied, while the cases of the simultaneous occurrence of 

 the two kinds of adventitious avicularia are innumerable. 

 Without going so far as to say that intermediate con- 

 ditions do not occur — a generalisation that could only be 

 established by very prolonged study — it may certainly be 

 maintained that it is the general rule for an avicularium 

 to assume one of the two types. In a suitable prepara- 

 tion it is usually quite easy to sort all the mandibles into 

 their proper group at first sight, and without having to 

 pause to consider doubtful cases. This fact is surely 

 significant, and it can at least be argued that in the blasto- 

 genic processes by which the avicularia have been developed 

 some differentiation or segregation must occur by which 

 the two kinds are constituted. If this differentiation should 

 prove to be analogous to the segregation which occurs 

 during the formation of gametes we should be able to 

 account for much that is at present perplexing in the 

 polymorphism of the Cheilostomata. We should in pai 



