July 8, 1909J 



NATURE 



57 



air were determined by comparing tlie rates of flow of 

 heat through them with the rate of flow through air when 

 the plates were at a fixed distance apart. The paper con- 

 cludes with a calculation of the " radiation constant," from 

 a determination of the absorption coefficient of the surfaces 

 of the plates when painted black, and the radiation R. — 

 The possible ancestors of the horses living under domestica- 

 tion, part i. : Dr. J. C. Evvart. By some naturalists it 

 is believed that domestic horses are the descendants of a 

 Pleistocene species {Equus fossilis) — now represented by the 

 wild horse (E. prsewalskii) of Mongolia — by others, the 

 liorses living under domestication are said to be a blend 

 of a coarse-headed northern species- allied to Prejvalsky's 

 horse, and a fine-limbed southern species which in pre- 

 historic times inhabited North Africa, or a blend of^ a 

 Prejvalsky-like northern species and a southern species 

 closeb- allied to E. sivalcnsis of the Pliocene deposits of 

 India! The examination of the skull, teeth, and limb 

 bones of horses found at Roman settlements and in the 

 vicinity of pile-dwellings indicates that domestic horses 

 originally belonged to several distinct types, viz. (i) a 

 type characterised by long limbs, by a long face, broad 

 and convex between the orbits, and strongly deflected on 

 the cranium, and by the crown of the fourth premolar 

 being from before backwards about 2-5 times the length 

 of the grinding surface of its "pillar"; (2) a type with 

 slender limbs, a fine, narrow, slightly deflected face, and 

 the crown of the fourth premolar about three times the 

 length of its " pillar " ; (3) a type with fairly slender 

 limbs, a long, narrow, somewhat deflected face, and the 

 crown of the fourth premolar about twice the length of 

 its "pillar"; (4I a type characterised by short, broad 

 metacarpals, a short face, broad and flat between the 

 orbits, and nearly in a line with the cranium, and by the 

 crown of the fourth premolar being twice the length of 

 its " pillar "; and (5) a type with short, wide metacarpals, 

 the face long and strongly deflected, and the crown of the 

 fourth premolar about 1-5 times the length of its " pillar." 

 Only the varieties characterised by molars with short 

 " pillars " are dealt with in this communication. The 

 possible ancestors of the short-pillared varieties are Equus 

 sivalcnsis of Indian Pliocene deposits, E. stenonis of the 

 Pliocene deposits of Europe and North Africa, and a new 

 species, E. gracilis. Arabs, barbs, thoroughbreds, and 

 other modern breeds with a long deflected face, broad and 

 prominent between the orbits, and the limbs slender, seem 

 to have mainly sprung from E. sivalcnsis, while certain 

 unimproved breeds with a deflected face, but very short 

 " pillars," are probably related to E. stenonis. E.xmoor, 

 Hebridean, Iceland, and other ponies of the " Celtic " type, 

 as well as ponies found in the south of France, the West 

 Indies, and Mexico, characterised by a fine narrow skull, 

 slender limbs, and the absence of ergots and hind chest- 

 nuts, are regarded as the descendants of E. gracilis, which 

 includes (i) the small species of the English drift described 

 by Owen as a fossil ass or zebra (Asintis fossilis) ; (2) the 

 small species of French Pliocene and Pleistocene deposits 

 known to palaeontologists as E. ligcris, and the small 

 species of North African Pleistocene deposits known as 

 E. asintis atlanticiis, and hitherto believed to be closely 

 related to, if not the ancestor of, zebras of the Burchell 

 type. By crossing experiments evidence has been obtained 

 of the wide distribution of horses of the E. gracilis type ; 

 that broad-browed Arabs and thoroughbreds, with the face 

 nearly in a line with the cranium, are mainly a blend of a 

 southern variety of E. gracilis (E. cahalhis lihyciis) and a 

 horse of the " forest " or .Solutr^ type, and that heavy 

 breeds have not inherited their coarse limbs from a species 

 closely allied to the wild horse of Mongolia. — The alcoholic 

 ferment of yeast-juice; part iv., the fermentation of 

 glucose, mannose, and fructose by yeast-juice : A. Harden 

 and W. J. Youngr. (i) Mannose behaves towards yeast- 

 juice, both in the presence and in the absence of added 

 phosphates, substantially in the same manner as glucose. 

 (2) Fructose resembles both glucose and mannose in its 

 behaviour, but in presence of phosphate is fermented much 

 more rapidly than these sugars, and the optimum concen- 

 tration of phosphate is much higher. (3) Fructose has 

 the property of inducing raoid fermentation in presence of 

 yeast-juice in solutions of glucose and mannose, containing 

 such an excess of phosphate that fermentation is only pro- 



XO. 2071, VOL. Si] 



ceeding very slowly. No similar property is possessed by 

 glucose or mannose. These properties of fructose indicate- 

 that this sugar when added to yeast-juice does not act 

 merely as a substrate to be fermented, but bears some 

 specific relation to the fermenting complex. All the facts 

 are consistent with the supposition that fructose actually 

 forms a part of the fermenting complex. When the con- 

 centration of this sugar is increased, a greater quantity 

 of the complex would be formed, and, as the result of this 

 increase in the concentration of the active catalytic agent, 

 the juice would become capable of bringing about the re- 

 action with sugar in presence of phosphate at a higher 

 rate, and at the same time the optimum concentration of 

 phosphate would become greater, exactly as is observed. 

 —The electrical reactions of certain bacteria applied to the 

 detection of tubercle bacilli in urine by means of an electric 

 current : C. Russ. The aim of these experiments was to 

 ascertain whether bacteria suspended in an electrolyte are 

 transmitted during electrolysis to either electrode, with the 

 view of the recovery of pathogenic bacteria from a patho- 

 logical fluid by such means. During electrolysis of certain 

 salts in which bacteria were suspended, the organisms 

 were found to migrate to one electrode ; in some instances 

 there was no migration. The effect was noticed to occur 

 with killed as well as with living bacteria. By testing 

 certain organisms in the same (but a small) series ot 

 electrolytes some differences of effect were found, though 

 this line of inquiry was not pursued. To utilise this 

 bacterial movement, an electrolyte in which tubercle baciUi 

 had shown marked kathodic aggregation was added to 

 tuberculous urine, and the kathode arranged in the form 

 of a bacterial trap. After electrolysis tubercle bacilli 

 entered the trap, which was eventually withdrawn, and the 

 organisms recognised in a stained film prepared from its 

 contents. A series of such urines was tested in this way, 

 and in each case tubercle bacilli were found in the trap. 

 In the final experiment a number of tubercle bacilli 

 (estimated at 500) were added to 100 c.c. normal urine, and 

 their detection attempted by separate investigators by 

 means of the centrifuge and current. By the centrifuge 

 none were found, while the current recovered 128 bacilli. 

 The results of this preliminary investigation may be sum- 

 marised as follows :— Certain bacteria under the influence 

 of a suitable current aggregate at one or other electrode. 

 The aggregation varies with the nature of the electrolyte 

 and is probably due to aflinity between the products o 

 electrolysis and the bacteria. It occurs with killed as \yeU 

 as with living bacteria. The aggregation by electrical 

 currents affords a means of collection and examination. 

 The differences in behaviour of various bacteria are such 

 as to suggest the possibility of utilising the method for 

 purposes of specific discrimination, but in this particular 

 The data hitherto obtained are not sufficient to warrant 

 definite statements.— The effect of the injection of the intra- 

 cellular constituents of bacteria (bacterial endotoxins) on 

 the opsonising action of the serum of healthy rabbits : Ur. 

 R Tanner Hewlett. In this investigation the effect ot 

 the endotoxins of the Bacillus typhosus. Micrococcus 

 Pyogenes aureus, and B. tuberculosis on the opsonising 

 action of the serum of normal rabbits has been studied. 

 The endotoxins were prepared by the Macfadyen process, 

 the rabbits were inoculated subcutaneously, and the speci- 

 mens for counting the number of bacteria ingested by 

 leucocytes were prepared in the usual manner. Human 

 leucocvtes were employed, and the counts were made on 

 fifty cells, (a) Typhoid Endotoxin.— The results for this 

 endotoxin are approximate only, as agglutination and 

 bacteriolysis are complicating f.actors. The aniount ot 

 endotoxin injected was o-i mgrm., prepared from an 

 avirulent strain. One day after injection a decided negative 

 phase had developed (opsonic index about o-2l, two days 

 after injection the index was rising (1-4). and /"^'"A?, '^ 

 maximum on the third day (3-3), after which it fell. Dilu- 

 tion of the serum to i in .; and i in 10 tended to increase 

 phacfocytosis. (b) Staphylococcus Endotoxin.— Bniotoxm 

 prepared from an old laboratory strain in a dose of 

 01 mgrm. produced a rise in the opsonic index to i-b, 

 which persisted for some weeks. Endotoxin (01 mgrm.) 

 prepared from a recently isolated strain produced a riseto 

 2-s. An equivalent 'dose of staphylococcus vaccine 

 (1000x10" cocci) produced a rise of the opsonic index to 



