August 24, 1882] 



NATURE 



407 



in the hypothesis previously quoted that the primitive streak is 

 the relic of a blastopore. 



At this time Balfour was working hard at his text-book of 

 " Comparative Embryology." His published papers were no less 

 numerous than before, but consisted in part of extracts from the 

 more speculative chapters of the forthcoming book. He, however, 

 published a paper, (Quarterly "Journal of Microscopical Science, 

 vol. xx., 1S80. ) containing the results of work scattered over two 

 years, on the. development of Spiders. He also published a 

 paper (Proceedings of the Zoological Society, 1881.) on the 

 skeleton of the paired fins, based upo 1 his work on Ela-mo- 

 branchs. In this he contests thj views of Gegenbauer and 

 Huxley, th t the primitive fin consists of a central muhi segmented 

 axis with many lateral rays, and is most nearly retained in 

 Ceratodus ; he rather considers the primitive form to be a 

 longitudinal bar running along the base of the fin (basipterygium), 

 and giving off at right angles a series of rays which pass into the 

 fin. He adheres to the view expressed in the " Elasmobranch 

 Fishes," (p. 101.) that the vertebrate limbs are remnants of two 

 continuous lateral fins. 



Another imp irtant paper of the same year dealt with the 

 placenta. Balfour supposed that in the primitive PlacentalU, 

 simple foetal villi, like those of the pig, projected from the 

 discoidal allantoic region of the chorion into uterine crypts. 

 The deciduate discoidal placenta of Rodents and Insectivores 

 is the first stage in advance of this primitive type. Then along 

 different lines diverge the zonary placenta of Carnivora, and the 

 diffuse placenta of Suida;, Lernuridae, &c. ; and the latter 

 becomes contracted down to the di coidal placenta of man, a 

 form in no way to be confounded with the prim tive discoidal 

 placenta of Rodents. 



He engaged also, in conjunction with Mr. W. N. Parker, 

 in a very important research; to be published in full in the 

 "Philosophical Transacti ins," on the "Structure and Develop- 

 ment of Lepidosteus." This paper contains an immense amount 

 of new matter, both anatomically and enbryological, and shows 

 that I.epidosteus, though a 'true ganoid, has very marked tele- 

 ostean affinities. 



Balfour's last published paper, (Quarterly Journal of Micro- 

 scopical Science, vol. xxii., 1882.) which appeared during his 

 recent illness, was written with the as istance of Mr. Deighton, 

 and related to the germinal layers of the chick. This paper 

 describes, in a very beautiful way, the djuble origin of the 

 mesoblast, partly from an axial strip of epiblast in the line of 

 the primitive streak, ani partly as two lateral plates differentiated 

 from the hypoblast in front of the primitive streak. 



Before his last, fatal journey, Mr. Balfour was engaged in 

 preparing a new edition of the "Elements of Embryology," 

 and in producing a very elaborate memoir on the "Anatomy 

 and Development of Peripatus." He had previously investi- 

 gated that animal, in 1879, and had cleared up the matter of its 

 segemental organs (overlooked by Moseley), and demonstrated 

 the presence of ganglia on its ventral nerve-cords. 



Mr. Balfour became a member of this Association in 1871, 

 the year after he entered Trinity College. At the brilliant 

 Belfast meeting in 1874 he read his first paper before the 

 Association on Elasmobranch Fishes ; and this paper and 

 Balfour's share in the keen discussion which followed are still 

 remembered with admiration by many. In 1880, at Swansea, 

 lie delivered an address, as Chairman of the sub-section of 

 Anatomy and Physiology, dealing with the mutual services 

 rendered by the evolution theory to embryology, and by embryo- 

 logy t) the evolution theory, with special reference to the 

 developmental history of the nervous system. In 1881, he was 

 appointed one of the two General Secretaries. 



But the great text-book of comparative embryology (" Com- 

 parative Embryology," vol. i., 1880, vol. ii. 18S1) is the real 

 monument of Balfour's fame. It is impossible to convey an 

 idea of the merits of this book. It grappled with the enor- 

 mous mass of scattered literature upon the subject, and formed 

 it all into a consecutive account, clear and accurate. Discordant 

 statements were weighed and estimated, frequently brought 

 into harm my by an ingenious explanation or by a new and 

 crucial observation. Countless investigations were repeated 

 and verified, and countless sugge-tions of important work, that 

 still remain- to be done, make the book as valuable to the 

 savant as to the student. Among the chapters (" Comparative 

 Embryology," vol, ii., chap. xi. xii. xiii.) most remarkable fir 

 broad and philosophic generalisations are those dealing with the 

 "Ance tral Form of the Chordata," "Larval Forms," and the 



"Origin and Homologies of the Germinal Layers." Balfour 

 accepts the gastrula as a stage in the evolu'ion of the metazoa, 

 and leans somewhat to invagination, as the more primitive pro- 

 cess than delamination i 1 the production of the gastrula. He 

 show- distinctly that the mesoblast arose in the fir.-t instance, 

 not independently, but as a differentation from the other two 

 layers, and that the mesoblast is a homologous structure 

 through nit the triploblastic metazoa. In the chapter on 

 " Larval Forms " he gives numerous reasons and arguments for 

 a larval development repeating the ancestral history, belter and 

 more fully than a foe'al development ; he reviews the types of 

 larvae (discriminating six types), the cases tending to produce 

 secondary changes in larvae, and suggests, as a hypothesis for 

 the passage from the radial to the bilateral type, that in a 

 pilidium-ilke larva the oral face elongated unequally, an 

 anterior part forming a prce-oral lobe, and a posterior outgrowth 

 the trunk, while the aboral surface became the dorsal surface. 

 He suggests that adult Echinodermata have retained, and not 

 secondarily acquired, their radial symmetry, and considers a 

 radially symmetrical organism, like a medu-a, as the prototype 

 of all the larval forms above the ccelenterates. Balfour does 

 not admit the specially close relatim-hip of the Chordata with 

 the Chretopods, which Dohrn and Semper maintain ; but con- 

 siders that the Chordata descended from a stock of segmented 

 worms derived from the same unsegmented types as the 

 Chretopois, but in which two lateral nerve-cords like those of 

 the nemertines coalesced dorsally instead of ventrally. He con- 

 siders that the mouth in ancestral Chordata was suctorial, and 

 was not formed, as Dohrn supposes, by the coalescence of two 

 visceral clefts. Finally, Balfour draws up a scheme of the 

 phylogeny of the Chordata, according to which the hypo- 

 thetical protochordata, with a notochord with a suctorial mouth 

 and very numerous gill-slits, acquired one by one, vertebrae, jaws, 

 an air-bladder, a pentadactyl limb, an amnion : each new acces- 

 sion characterising a hypothetical protogroup, from which some 

 existing group is supposed to have diverged. 



Tho-e of my hearers who had not followed Balf >ur, scientific 

 labours, but who merely knew him as one of the most respected 

 workers in the field of biology, will I trust, even from my brief 

 sketch, have formed some idea of the activity and originality of 

 his mind, and will understand how his death has occasioned a 

 feeling almost akin to despair, in that he occupied a place which 

 it appears to us now impossible to fill. " How are the mighty 

 fallen, and the weapons of war perished ! " 



On the Growth of our Knowledge of the Process of 

 Secretion in the Animal Kingdom. 



The Vicivs of the Ancients. — It was known to the ancients 

 that organs of the body exist which are concerned in the 

 separation from it of excrementitious substance--, although the 

 greatest doubts prevai'ed as to the organs to which such functions 

 should be ascribed. Thus we find Hippocrates defining it as 

 characteristic of glands that they occur in moist parts of the 

 body ; but showing his ignorance of the true relations of glands 

 to secretion by connecting them with the formation of hairs, and 

 discussing the question which we find our own Wharton debating 

 again in the seventeenth century, and which he formulates, 

 " Num cerebrum ad glandularum numerum vel viscerum 

 accedat." The general opinion of the ancients, and the opinion 

 which was adopted and by Galen, was that the glands were 

 sieves or collanders (coli), which served to strain off from the 

 blood purely excrementitious substances. The liver and kidneys 

 were strangely enough removed from the group of glands and 

 placed amongst the viscera. The first writer who appears 

 systematically to have treated of the glands was the before- 

 mentioned Wharton in his " Adenographia sive glandularum 

 totius corporis descripti >." Although this author certainly added 

 to the existing knowledge of the descriptive anatony of secreting, 

 organs, his views on the functions of glands were strangely 

 fanciful and erroneous. 



The gUnds he considered to be especially related to the 

 nervous system, the viscera, so-called, to the blood-vessels ; such 

 glands as the pancreas, and the salivary and lachrymal glands 

 being engaged in separating excrementitious substances from the 

 nervous system. It was in 1665 that the great anatomist 

 Mai; ighi(" Exercitatio Anatomica de Renibus") first attempted 

 to investigate the structure of glands in a truly scientific spirit, 

 endeavouring to establish a relationship between simple glaadu- 

 lar follicles and such complex glands as the liver. All glands he 

 believed to contain as ultimate elements bodies which he termed 



