456 



NATURE 



[September io, 1903 



which the embryologist meets with in the segmentation of 

 the egg suggest that there are considerable differences in 

 these respects between the eggs laid by a single parent in 

 a single act of oviposition. Moreover, the manner in which 

 the young eggs of the insects are nourished in the tubular 

 oviduct before they are ready for fertilisation gives very 

 little support to the view that the amount of yolk deposited 

 in each c^^ is identical. 



The second consideration under this heading is possibly 

 of even greater importance. Vernon ' has shown that the 

 size and other characters of echinoderm larvae vary very 

 considerably according to the freshness or staleness of the 

 conjugating ova and spermatozoa. For example, he found 

 that when the fresh spermatozoa of Strongylocentrotus 

 fertilised the eggs which had been kept eighteen hours of 

 the same animal, the larvae differed from the normal larvae, 

 — 17-6 in body length and —15 per cent, in arm length, 

 and when the fresh eggs were fertilised by spermatozoa which 

 had been kept eighteen hours the resulting larvae differed 

 from the normal by +11 per cent, in body length and by 

 — 32-8 per cent, in arm length. 



This consideration is practically eliminated in the case 

 of the worker-bees by parthenogenesis, but it cannot be set 

 aside in the case of the drones nor in the cases of the broods 

 of other animals which do not exhibit the phenomenon of 

 parthenogenesis. A comparison of the curve of variation 

 of some character, common to both, in drones and worker- 

 bees from one hive would perhaps throw some light on the 

 general importance of this character. 



Before leaving this part of the subject, I must call atten- 

 tion to two results bearing upon it, obtained by De Vries 

 in his botanical investigations, and related by him in his 

 very important work entitled " Die Mutationstheorie." 

 This observer found that the younger the seedling is the 

 greater is the influence of external circumstances upon its 

 adult characters, and in the second place that an even 

 greater influence is exerted upon the characters of a plant 

 by the external circumstances affecting the mother-plant. 

 If these results hold good for animals as they do for plants, 

 we should expect to find, then, that the external circum- 

 stances affecting the mother at the time she is maturing the 

 eggs in her ovaries and the external circumstances affecting 

 the embryo before and during the larval period are of far 

 greater importance in affecting the curve of variation of 

 the adults than are the external circumstances affecting the 

 young in their period of adolescence. We must come to the 

 conclusion, from these considerations, that the general 

 variability of a brood or progeny of a single pair of parents 

 must be very largely the effect of the varying conditions 

 affecting the gametes from the earliest stages of their 

 genesis in the gonophore, the fertilised ovum, and the early 

 stages of development. We find, however, as I have already 

 pointed out, that some characters are much more influenced 

 by external circumstances than others. Weight and stature 

 in human beings, for example, are probably much more in- 

 fluenced than the colour of the iris or the shape of the 

 fingers. We may, indeed, recognise two kinds of characters, 

 connected, of course, by a complete series of intermediate 

 links, which may be called, for convenience sake, plastic 

 characters and rigid characters. 



Now, in some animals, the characters that are rigid are 

 much more numerous than they are in others. For example, 

 adult salmon or perch are much more variable in size and 

 weight than adult herrings or mackerel ; some species of 

 butterflies are much more variable in the colour and pattern 

 of their wings than other species ; some species of birds are 

 much more variable in their plumage than others are. 

 Several other examples could be chosen to illustrate this 

 point from the higher groups of animals ; but I wish 

 particularly to call your attention to several instances found 

 in the Coelenterata, because it was the special study of this 

 group of animals that led to the train of thought I have 

 ventured to put before you. 



In all the sedentary forms of Coelenterates the mouth 

 is surrounded by a circlet of tentacles. These organs are 

 used for catching and paralysing the prey and passing it 

 to the mouth to be swallowed. They are also very delicate, 

 and indeed the only specialised organs of sense performing 

 a function similar to that of the feelers or antennae of Arthro- 



1 H. M. Vernon : " The Relations between the Hybrid and Parent-forms 

 of Echinoid Larvx." Phil. Trans. 1898, B. p. 465. 



NO. 1767, VOL. 68] 



poda. There can be no exaggeration in saying, therefore, 

 that they are of the utmost importance to the animal. In 

 some groups of Coelenterata, however, we find that they 

 are fixed in number, but in others that they are variable. 



In the Alcyonaria, for example, the number of tentacles 

 of the adult polyp is eight. I have examined many 

 thousands of polyps belonging to the suborders Stolonifera, 

 Alcyonacea, Gorgonacea, and Pennatulacea, and I have not 

 found a single example of an adult polyp with either more 

 or less than eight tentacles. This is a character, then, 

 which is remarkably well fixed in the Alcyonaria. It does 

 not fluctuate at all. The tentacles of the Hydrozoa, and of 

 many of the Zoantharia, on the other hand, fluctuate con- 

 siderably in number. In some forms, such as Tubularia 

 among the Hydroids, and Actinia among the Zoantharia, 

 the number of tentacles is considerable, and it is not, 

 perhaps, surprising to find variations in their number. But 

 in many cases, when the number of tentacles is small, there 

 is also frequent variation. In Hydra viridis, for example, 

 the number of the tentacles is 6, 7, or 8, and more rarely 

 5 or 9. 



Again, in the Alcyonaria, the number of mesenteries of 

 the adult polyp is always eight ; never more and never less. 



In the Zoantharia, on the other hand, the number varies 

 not only in different suborders and families, but even in 

 different individuals of the same species from a single 

 locality. Parker found, for example, that the number of 

 non-directive mesenteries in the sea-anemone Metridium 

 marginatum, collected at Newport, R.I., varied from four 

 to ten pairs in those forms with the normal number (2) of 

 directive mesenteries, and that there were further variations 

 in the number of non-directive mesenteries in those forms 

 with an abnormal number of directive mesenteries. In 

 fact, of the 131 adult specimens collected, only 40 or about 

 33 per cent, exhibited the arrangement of mesenteries which 

 is regarded as normal for the species. On the other hand, 

 Clubb found that of the specimens of another common sea- 

 anemone. Actinia equina, only 424 per cent, showed vari- 

 ations from the normal mesenterial arrangement for the 

 species. We have then, in these examples, a set of organs 

 which are very variable in one genus (Metridium), much 

 less variable in another (Actinia), and perfectly fixed or 

 rigid in another series of genera (the Alcyonaria). 



Passing on, now, to the character " shape." Not many 

 years ago the systematic zoologists, who directed their 

 attention to the sedentary Coelenterates, based their specific 

 diagnoses very largely on the shape of the colonies. Thus 

 we have introduced such names as Millepora alcicornis, 

 M . ramosa, M. plicata, Madrepora cervicornis, M. prolifera, 

 M. palmata, Alcyonium digitatum, A. palmatum, &c. 

 Zoologists are now agreed, however, that the shape of these 

 colonies is so variable that in most genera it is of very 

 little value for the separation of species. In fact, I have 

 elsewhere given reasons for holding the view that the widely 

 distributed and very variable genus Millepora is represented 

 by only one true species. But what is true for most 

 sedentary Coelenterates is not true for all colonial 

 Coelenterates. In most of the genera and species of 

 Pennatulida, for instance, the shape of any one individual 

 of a species is almost identical with that of any other. A 

 Funiculina quadrangularis, from the west coast of Scotland, 

 is similar in shape to one of the same species from the coast 

 of Norway. A Pennatula murrayi, from the reefs of 

 Funafuti, is similar in shape to one from Ceram. In other 

 words, the character " shape " is extremely plastic in 

 Millepora and Madrepora, but very slightly plastic or almost 

 rigid in Pennatula and Funiculina. 



This difference in the plasticity of the character "shape " 

 in Millepora and the Pennatulids must be associated with 

 the fact that the young Millepora colony is unable to move 

 from the spot where the larva settles, whereas the Penna- 

 tulid is capable of moving from place to place throughout 

 life. The Millepora colony must either accommodate itself 

 to the environment in which it begins life or perish, but 

 the young Pennatulid can, within certain limits, travel to 

 the environment that suits itself. 



The shape of a growing coral or sedentary Alcyonarian 

 on a reef must accommodate itself to the depth of water, 

 the position of neighbouring zoophytes to itself, the direc- 

 tion of the tides, and other influences ; and such a power of 

 accommodation is essential for the species in the struggle 



