340 



NATURE 



\Atigust 7, 1879 



I 



guished by the fineness and abundance of the minute 

 pores by which the surfaces of their skeletons are pierced. 

 The animals belonging to these corals had not been ex- 

 amined until they were investigated by Agassiz, and as 

 he was not able to make any extended investigation of 

 their structure, his results were long accepted by many 

 naturalists with considerable hesitation. During the 

 voyage of H.M.S. Challenger I studied the structure of 

 several species of Millepora in detail, with the result of 

 confirming Prof. Agassiz's results and yielding a detailed 

 account of the minute structure of these organisms, which 

 is almost complete, excepting with regard to their gene- 

 rative organs, which remain as yet entirely unknown. 



In pursuing my observations on corals, I discovered 

 that another family of stony corals, as well as the Mille- 

 porids, is also Hydroid in structure. This is the family 

 of the Stylasterida:, an account of the structure of which 

 was selected by the Royal Society as the Croonian lec- 

 ture for last year, and has just appeared in the new 

 volume of the Philosophical Transactions. In the present 

 article I have brought together the principal results of 

 interest which are stated in detail in the lecture. 



One of the Stylasteridas (Allopora) had previously been 

 examined in the living condition by Prof. G. O. Sars, and 

 Sars had suspected that this coral might be Hydroid like 

 the IMilleporidx, but he had been unable to work out the 

 details of the structure of the organism and to prove the 

 matter with certainty. Several observers, the late Dr. 

 Gray, Prof. V^errill, and the Count de Pourtales, had ob- 

 served the distinctness of the Stylasterida;, and noticed 

 that there were remarkable peculiarities characterising 

 this family of corals. 



The Hydroid stony corals, the Milleporidas and Stylas- 

 teridas, I have placed in a special sub-order, the Hydro- 

 corallinffi. Though the two families are well distinguished 

 from one another, they show many close resemblances in 

 structure. 



Amongst the Hydrocorallinx there do not exist, as 

 there do amongst the Anthozoan corals, any simple or 

 solitary species ; there are no Hydroid corals comparable 

 thus to the mushroom corals, the only forms known are 

 compound colonies. In the case of the Anthozoan corals 

 it seems probable that in the progress of development a 

 simple ancestral form derived from a sea-anemone deve- 

 loped a calcareous skeleton, and that from this solitary 

 form compound corals were derived as subseciuent modi- 

 fications ; or, rather, it is not unUkely that several solitary 

 ancestors developed calcareous skeletons independently, 

 and that from each of them different compound forms 

 resulted. In the case of the Hydrocorallinas, on the other 

 hand, it seems probable that the calcareous skeleton was 

 first developed as a support to already formed colonies, 

 and that no solitary ancestor with a calcareous support 

 preceded them. 



Almost all the recent Anthozoan corals belong to the 

 Hexactinia or corals which, like the common sea ane- 

 mones, have the radially disposed soft structures of their 

 bodies and the corresponding radial plates of their 

 skeletons arranged in multiples of the number six. 

 These Hexactinian Anthozoan corals are termed the 

 Madreporaria. It is a very remarkable f ct that amongst 

 all the vast number of species of compo 1 d Madreporaria 

 known, there seems to exist no instance of a modification 

 of certain of the animals composing the colonies by a sub- 

 division of labour amongst them for the general good to 

 the colony. Amongst Alcyonarian corals and Hydroid 

 corals such a sub-division of labour exists, but for some 

 reason or other such high specialisation seems never to 

 have been attained amongst Madreporaria. 



In the case of the Hydrocorallina;, the subdivision of 

 labour amongst the members of the colonies is carried to 

 a most interesting perfection. It reaches considerable 

 completeness in the Milleporidas, but in the more ad- 

 vanced Stylasterida; attains a most elaborate complexity. 



In all the Hydrocorallince the hard skeleton is very 

 porous, being traversed in all directions by canals which 

 branch and join one another in all directions. Within these 

 canals are lodged corresponding branching and anasto- 

 mosing tubes composed of soft tissues, which form a 

 complex meshwork within the coral mass, and convey a 

 general circulation common to all the members of the 

 colony. In all the Hydrocorallinte two kinds of polyps or 

 zooids occur. The more numerous kind are devoid of 

 any mouth or stomach, and act simply as catchers of food 

 for the colony. These are hence termed dactylozooids. 

 The less numerous kind have each a mouth and stomach, 

 and are hence termed gastrozooids ; they receive the food 

 from the dactylozooids, and swallow it, and their bases 

 being in direct communication with]the general circulation, 

 they nourish with the results of their. digestion the dactylo- 

 zooids and all the component parts of their colony. 



In the MiUeporidK the dactylozooids, when expanded, 

 are long and slender, and are provided all along their 

 lengths with short tentacles each of which bears a knob 

 at its end. In the case of the Stylasterida;, however, the 

 dactylozooids have quite lost their tentacles, and are 

 simply long, slender, tapering bodies, reduced to the 

 aspect of simple tentacles themselves. In the Mille- 

 poridas the gastrozooids are provided with short tentacles 

 round their mouths, and such tentacles are also present 

 in the case of the gastrozooids of most of the genera of 

 Stylasterida;, the number present varying in the case of 

 each genus. In some genera of Stylasteridas, however, 

 even the gastrozooids have lost their tentacles, and re- 

 main as simple cylindrical digestive sacs ; in these in- 

 stances the entire colonies are devoid of tentacles 

 altogether. 



The zooids arc lodged within pits or pores on the 

 surfaces of the corals, which are termed gastropores or 

 dactylopores, according to the kind of zooid belonging 

 to them. The gastropores are larger than the dactylo- 

 pores in correspondence with the size of the respective 

 zooids. In most species of Millepora and several genera 

 of Stylasterida; the gastropores and dactylopores, inter- 

 mingled with one another, are scattered irregularly all 

 over the surfaces of the corals ; but in one species of 

 Millepora occurring at Tahiti the pores are gathered with 

 some considerable regularity into circular groups, each 

 of which is composed of a single centrally-placed gastro- 

 pore and a surrounding zone of six or seven dactylopores. 

 In this case the zone of dactylozooids in each group or 

 system ministers to the wants of the single gastrozooid 

 of that system. ^1 



The complexity of relations of the zooids advances*' 

 no further in the case of the Milleporida:, but in 

 that of the Stylasterida; many additional complications 

 exist. In several genera (/i//c/o;-rt, Stylasier, Cryptohclia, 

 Astylus) the pores occur only in regular circular systems, 

 which are termed cyclosystems. In each of these cyclo- 

 systems there is a deep centrally-placed gastropore and 

 a circular zone of from five to upwards of twenty dactylo- 

 pores. The mouAs of the dactylopores, instead of being 

 circular in outline, are drawn out into the forms of long 

 sHts, which are disposed in the cyclosystems in a regu- 

 lar radial manner towards the central gastropore. In 

 species in which the dactylopores are numerous and 

 closely packed in the systems, a thin wall only of hard 

 coral skeleton is left intervening between them. Hence 

 each system has closely the appearance of an ordinary 

 Anthozoon coral cup, with its radiating septal plates, and 

 so close is the resemblance that Gray and all earlier 

 observers did not doubt that Stylaster and its allies were 

 essentially similar in structure to the ordinary branching 

 corals, such as Oculina. 



The essential difference between a cyclosystem of a 

 Stylasterid and an Anthozoan coral cup is, however, as 

 wide as possible. The radiating plates in the case of 

 the Anthozoan coral are skeletal structures developed 



I 



