January 8, 1920] 



NATURE 



481 



in such a stdel by slow cooling from the tempering 

 temperature, and is revealed by a low absorption of 

 ■energy in the single-blow impact test on notched bars. 

 The author has found that wide differences in the 

 impact figure with almost identical tensile test results 

 can be produced by suitable heat treatment. He has 

 also found that whereas, after hardening, every tem- 

 pering treatment involving a final rapid cooling from 

 600° C. or above produced good impact figures, a 

 final slow cooling produced a considerably lower, and 

 often a bad, impact figure; further, that in any given 

 steel the degree of brittleness which can be produced 

 by a given condition of tempering depends t.n the 

 original hardening temperature. The higher this is, 

 the lower is the impact figure. He has also found 

 that reheating to about 520° C. produces brittleness, 

 whatever the subsequent rate of cooling, and that 

 this can be removed by reheating to between (100° C. 

 and 670° C. and cooling rapidly. 



These results can be explained on the assumption 

 that a critical temperature or temperature range exists 

 in the neighbourhood of 550° C, above which the 

 tough, and below which the brittle, condition 

 IS stable. Quick cooling through this tempera- 

 ture retards this change, and the unstable tough 

 condition is retained. Slow cooling results in the 

 production of the stable brittle condition. If the 

 tough material is heated to a temperature rather below 

 the change point, the rate of change to the brittle 

 condition is at a maximum, and brittleness results. 

 The rate diminishes rapidly with fall of the tem- 

 perature, and below 450° is negligible. Provided, there- 

 fore, the critical temperature is not exceeded, the rate 

 of cooling after this reheating is immaterial. Mr. 

 Greaves does not show anv cooling or heating curves 

 of his steel. Those published by Messrs. ."Xndrew, 

 Greenwood, and Green on a steel of approximatelv 

 the same composition indicate that the carbon-change 

 point on cooling occurs between about 4qo° and 465° 

 C. The character of the curve obtained depends 

 upon the initial temperature from which the steel is 

 cooled. H. C. H. C. 



CEPHALODISCUS AND THE 

 ARCmCHORDATES.^ 



THE history of Cephalodiscus, dredged at 

 245 fathoms in the Strait of Magellan by the 

 Challenger, and at first taken for an Alga, and then 

 for a compound .Ascidian, goes back only three dozen 

 vears. Moreover, the sole species (C. dodecalophus) 

 held the field for tuenty-one years before the other 

 species made their appearance ; but now, with Dr. 

 Ridewood's memoir before us, the total number of 

 species reaches from twelve to sixteen, though further 

 research may reduce that number. It is noteworthy 

 that whilst the majority group themselves around the 

 South Pole, four occur in the Indian and Pacific 

 Oceans. 



In the present memoir Dr. Ridewood, already 

 known as an authority on the subject, keeps to the 

 classification adopted previously, the group Ptero- 

 branchia ^^spidophora of .\IIman) having three sub- 

 genera of Cephalodiscus, viz. Demothecia. colony 

 branched, with a continuous cavity throughout the 

 roenoecium ; Idiothecia, colony branched, but each 

 aperture leading into a tube occupied bv one zoold 

 and its buds; and Orthothecus, in which the colony is 

 cake- or cone-like, each aperture entering a tube 

 holding a zooid and its buds. The author first treats 



I " Bri'i^h Antarctic {Terra Km'a) F.xpfdition. igio. Ceph.-i'odi^cu';. 

 Ry Dr. W G. Ridewood. With 12 texr-fienres, 5 pLitffs. and a map 

 (Published by the Trustees of the Unt'sh Museum (Natural History), 1918 ) 

 Price I2J. , 



NO. 2619, VOL. 104] 



of the structure of the zooids, the similarity of which 

 throughout the whole series is noteworthy ; only in 

 the reduced male zooids of C. sibogae, Harmer, is 

 there a divergence. This fact alone would give differ- 

 ences due to variations in the coenoecium less weight. 



Amongst other features of interest are the enlarge- 

 ments at the ends of the tentacles, for instance, in the 

 original species, which the author terms " end- 

 swellings with refractive beads," and it is curious that 

 no special function has been assigned to them. Similar 

 enlargements at the tips of the branchial filaments are 

 prominent features in Filograna and the so-called 

 Salmacina, and great weight has been placed on 

 them .specifically, and even generically, by certain 

 observers. In all probability they are sense-organs 

 in both groups, since they are not connected with 

 secretion, nor do they perform the function of opercula 

 in Filograna (a Serpulid), in which form they are 

 present or absent with puzzling indifference, for the 

 plasticity of the species is phenomenal. The changes 

 in the character of the epithelium on the dorsal and 

 neural surfaces of the arms, and on the two surfaces 

 of the post-oral lamellae, are probably due, as in other 

 forms {e.g. the Serpulids), to differences in function. 

 The length of the testis is thought by the author to 

 be a specific character, but that of the ovary is not. 



Details are given of a new species, C. evansi, a 

 branched form, in which each ostium leads into a 

 tube ending blindly in the middle of the branch. The 

 other three species procured in the expedition were 

 formerly known, viz. C. nigrescens, Lankester, 

 C. densus, Andersson (which the author considers to 

 be a variety of the next), and C. hodgsoni, Ridewood. 

 Thereafter a discussion on the Demothecia occurs, the 

 species being extremely difficult to distinguish either by 

 coenoecium or zooids, and it is possible that future 

 observers may reduce the number of species, since the 

 variations of both coenoecium and zooids in a single 

 species are considerable. 



No new feacure is given in connection with repro- 

 duction and development further than that the author 

 thinks there is no certain relation between the number 

 of arms and the sex, as .Andersson did, and that in 

 C. hodgsoni the short stalk of the egg spreads over 

 the egg-shell. Males, females, and hermaphrodites 

 are found in C. hodgsoni, C. aequatus, C. nigrescens, 

 C. solidus, and C. densus, whilst no males have yet 

 been found in C. dodecalophus , C. levinseni, and 

 C. gracilis. In C. sibogae and C. agglutinans only 

 males are known, and in C. gilchristi and C. evansi 

 both sexes frequent the same colony. 



The author makes no allusion to the systematic 

 position of Cephalodiscus in zoological classification, 

 or to the homologies of the organs which have re- 

 ceived the attention of many zoologists in connection 

 with that classification. Dr. Masterman's Archi- 

 chordata (Trimetamera), therefore, stands as before, 

 with its two classes (i) Hemichordata {e.g. Balano- 

 glossus) and (2) Diplochordata (e.g. Phoronis, Cephalo- 

 discus, and Rhabdopleura), though not without dubiety 

 in certain aspects, which even the labours of Spengel, 

 Weldon, Cori, Fowler, De Selvs - Longchamps, 

 Lankester, Harmer, Gilchrist, Ridewood, Schepotieff, 

 Davidoff, Hill, Gravier, Pixell, and Roule have not 

 quite elucidated. Much of the dubiety is connected 

 with the notochord and the gill-slits. Dr. Harmer 

 thought that the proboscis-vesicle and " heart," with 

 the notochord, essentially agreed with the condition 

 in Balanoglossus, as described by Mr. Bateson ; but 

 Dr. Masterman, keenly working at Actinotrocha, 

 Tornaria, Phoronis, and the young forms of Cephalo- 

 discus, held that the primitive types had a double noto- 

 chord, and his beautiful and accurate drawings speak 

 for themselves whatever interpretation may be put 



