ded eat sped ty «' Cr) Rd 4A re “aod we Oe ened OH thet eh oe Gabe sab sw hai ed ett shed aon Wet Om dds Had att debs ede dew Atna n “ ait oh alte Molen ae 04 bee L (Aided Sy oh 6 het ” id rr ney write Memo a het eH OM dedi Tae y $24 rebated dite 6 tO a eon eee F gabe al A show thao a4 de aay 4 nd dete ing meee ee 44948 td ins err hh Pere 4 Sat ‘ +a h ad ? oo er oer) rise) heed’ Gat Ress atin Fictsi-a ate One detee ie en fehl ded Ah Neha 4a +6 cy eda esl wa Cry yer 254A = anaee Fe ey ore) bedded t* deaaed lke ae Che sae ‘ 16 ple eR ded tome beet AE ae dive en ete eG bed * S Cee ee Dodd i L a> owt 4 rary ' © Pease Y dest Medieg eM +0 Fide bende d Hee 9 lle ead ewes jo ot ead OR KDR SON aw aT eal-aeare 4M, dod Def ae tad ae rt mer IK ioe * Cag err rey “os Semele u 3 ei eee Ses Peril ea Pe Sete y= OA and Det Dt pee etd aS tie i+ heme aes Pi waoc' Sbsoie AE eared o) SG bs8N yi ain Toay * * Hee ad ae Ld 6A Gedoming Ae 9 9'9 Boked Seger Ye eaed ae 4 <4. ee Ck 4 eee ee 1 wa Hetew yO rws Mia Me OG SE edhe meth WE coe vy, TAR tod ee “4 « Cay eee ean Cee erro noe Wea jah arid we At" sibel matin sd Se Peery a 1m =~ Fy nr ; - "ory 7a = oan 4 ae a Wig ; ‘ . a AL 7 « » P \ xy i.® 1 a PROCEEDINGS OF THE ROYAL SOCIETY. Szcrion B.—BrotoagicaL ScrENcES. The Effect of Plant Growth and of Manures upon the Retention of Bases by the Soil. By A. D. Haut, M.A., and N. H. J. Mritugr, Ph.D. (Communicated by H. E. Armstrong, LL.D., Ph.D., F.R.S. (From the Lawes Agricultural Trust.) Received March 30,—Read May 11, 1905.) Introductory. The following investigation deals with the changes in the amount of calcium carbonate, the chief substance in the soil acting as a base, which are brought about by natural agencies, by manuring, and particularly by the growth of plants. Since Cavendish discovered that calcium carbonate dissolves in water charged with carbon dioxide, and ascertained the presence of calcium bicar- bonate in many natural waters, it has been recognised that the calcium carbonate present in most soils must be subject to regular loss. The air enclosed in the soil contains a considerable proportion of carbon dioxide derived from the roots of plants and decaying organic matter (Wollny, for example, found amounts varying from 3 to 14 per cent. by volume), and the . soil water, after attaining a state of equilibrium with the gas at this partial pressure, becomes an effective solvent of any calcium carbonate that may be present, removing it as bicarbonate into the drains or the general stock ot underground water. That such dissolution does take place in the upper layers of the soil is evident from the analyses made by A. Voelcker* and * © Journ. Chem. Soc.,’ 1871, p. 276. VOL, LXXVII.—B. B 2 Mr. A. D. Hall and Dr. N. H. J. Miller. —[ Mar. 30, Frankland* of the drainage waters from the Broadbalk wheatfield at Rothamsted.t Although the tile drains, from the flow of which the waters in question are derived, rarely lie more than 2 feet 6 inches below the surface of the soil, yet the drainage water from the unmanured plot contained on the average 99 parts per million of CaO, of which 84 were in the state of bicarbonate, and the water from the plot receiving farmyard manure every year contained 147 parts of CaO per million, of which 72 may be regarded as in the state of bicarbonate. Such concentrations, representing a hardness of about 17 and 26 degrees respectively, though by no means equal to those of truly calcareous waters, are above the average of natural waters in this country; yet, as will be seen later, the calcium carbonate from which they are derived is practically confined to the uppermost 9 inches of soil. While such losses may be regarded as natural, it has long been known that many of the substances applied to the soil under the general term of artificial manures react with the calcium carbonate there present and bring about its dissolution. Liebig, for example, pointed out that the di-hydrogen calcium phosphate (“superphosphate”) contained in bones made soluble by treatment with sulphuric acid, reacts at once with the bases in the soil and becomes again insoluble. The researches of Wayt and A. Voelcker§ showed further that the retention of ammonium and potassium salts by cultivated soils is always preceded by a double decomposition with calcium carbonate, the bases being retained as carbonates while the acids appear in the drainage waters combined with calcium. A. Voelcker’s analyses of the Rothamsted drainage waters,|| when dealing with the plots receiving salts of ammonium, potassium, etc., as manures, show the same reactions taking place on a large scale. Again, the process of nitrification, going on in all normal soils, requires some base to combine with the nitrous and nitric acids produced by the oxidation of the ammonia and other nitrogen compounds.§ In an ordinary way this base is supplied by calcium carbonate, hence a further source of loss to the calcium carbonate of cultivated soils. The soils of the Rothamsted experimental plots afford peculiar facilities for the study of the rate at which these losses of calcium carbonate, both natural * “Journ. Roy. Agri. Soc.,’ 2nd Series, vol. 18, 1882, p. 14. + See Table X. t ‘Journ. Roy. Agri. Soc.,’ 1st Series, vol. 11, 1850, p. 313, and vol. 13, 1852, p. 123. § ‘Journ. Roy. Agri. Soc.,’ 1st Series, vol. 21, 1860, p. 105, and vol. 25, 1864, p. 333. || Loc. cit. {7 ‘Instruction sur létablissement des Nitriéres, Paris, 1777; Warington, ‘Trans. Chem. Soc.,’ 1879, p. 429. 1905. ] On the Retention of Bases by- the Soil. 3 and induced by the application of saline manures, are taking place. In most cases the plots have received the same mauures, year after year, for more than fifty years, and though unfortunately samples of the soil were not taken at the starting of the experiments, yet in the case of the Broadbalk Field, on which wheat has been grown continuously since 1843, a set of samples drawn in 1856 has been preserved, in addition to samples drawn in 1865 and in 1893. From the Hoos Field (continuous barley since 1852) samples were drawn in 1868, 1882, and 1904—5, and from Agdel]l Field (under a four- course rotation since 1848) samples exist which were drawn in 1867, 1874, 1883, and 1905. Furthermore, the calcium carbonate in the Rothamsted soil is of extraneous origin, and is entirely localised in the surface layer which is stirred by the plough. The subsoil, from which the surface soil is undoubtedly derived, belongs to the drift formation of “ clay-with-flints,” characteristic of the chalk plateau, and consists of the débris of the chalk formation largely mixed with sands and clays of the Reading series.* It normally contains little or no calcium carbonate, although it is partly derived from the chalk formation and rests upon the solid chalk at a varying depth of 8 to 12 or 20 feet. In the eighteenth century, however, a characteristic feature of the agriculture of this district of Hertfordshire was to manure the land by sinking pits through the clay to the chalk, which was then lifted and spread in considerable quantities. Arthur Youngf quotes from Walker’s Survey of 1795—* the now prevailing practice of sinking pits for the purpose of chalking the surrounding land therefrom . . . The most experienced Hert- fordshire farmers agree that chalking of lands so circumstanced is the best mode of culture they are capable of receiving.” Evidence of the former prevalence of this practice of chalking may be seen by the existence in each of the Rothamsted fields of a “dell,” a depression representing the fallen-in pit from which the chalk was extracted. A certain rawness of the soil round the edges of these “dells” still bears witness to the disturbance ereated by the excavation, though it is known that nothing of the kind was done during the late Sir John Lawes’ possession of the estate, which dates back to 1834. Probably the pits were but little worked after the close of the eighteenth century, and certainly neither chalk nor lime has been applied to the plots since they were put under experiment. At the present time the * See H. B. Woodward, ‘“‘Report of the Soils and Subsoils of the Rothamsted Estate. Summary of Progress of the Geological Survey, 1903.” + “‘ Report on the Present State of the Agricuiture of Hertfordshire,” presented to the Board of Agriculture, 1804, B 2 4 Mr, A. D, Hall and Dr. N. H. J. Miller. [Mar. 30, chalk is visible only in the upper soil, and is there present in small rounded nodules varying in diameter from 3 or 4 mm. downwards. I.—CALCIUM CARBONATE IN ROTHAMSTED SOILS. A. Analytical, The first section of this paper deals with the determinations of the amounts of calcium carbonate present in the soils and subsoils of certain of the plots in the Broadbalk, Hoos, Agdell, and Little Hoos Fields, the samples having been drawn at the dates specified above and again in 1904—5. As the calcium carbonate is of artificial origin, and was probably distributed with considerable irregularity, however much this may have been equalised by the subsequent working to which the soil has been subjected, it cannot be expected that the samples analysed will represent the whole soil of the plots with the same degree of accuracy as would be attained in the case of some original constituent of the soil. Arable soils only are considered ; with soils in permanent grass the question is complicated by the well-known action of earthworms, which, as demon- strated by Darwin, bury the surface layer by constantly bringing fine subsoil to the top. The determinations of calcium carbonate have been made by means of an apparatus described by one of us in conjunction with Dr. E. J. Russell.* The results are calculated from the volume of carbon dioxide evolved on treating the fine soil with dilute sulphuric acid in vacuo, due provision being made to bring into account the carbon dioxide remaining dissolved in the reacting liquid. In a few cases, where the percentage of calcium carbonate was very low, the carbon dioxide evolved by treating a considerable quantity of the soil with acid was absorbed by caustic soda and determined by double titration. For these determinations we have to thank Mr. Arthur Amos, B.A. The amount of magnesium carbonate present is too small to affect the results, and in any case, as the real quantity sought is the amount of readily available base in the soil, it is desirable to express it always in the same terms. The soil samples were all taken in the same way: a steel frame 6 inches or 1 foot square and 9 inches deep is driven into the ground, and its contents are carefully picked out; this gives the soil proper. The surrounding soil is then dug away and the frame is driven down another 9 inches. The contents now represent the subsoil at the second depth of 10 to18 inches. The process is then repeated to as many holes as may be required. In this way samples are taken from four, six, or eight holes on each plot, according to its size. * Hall and Russell, ‘Trans. Chem. Soc.,’ vol. 81, 1902, p. 145. 1905. | On the Retention of Bases by the Soil. 5 The samples, after drying at a temperature not exceeding 60° C., are roughly powdered and put through a woven wire sieve with a mesh of } inch (this sieve passes a little more than the 3 mm. round-hole sieve now commonly used). From the fine earth thus obtained from each hole composite samples representing the whole plot at each depth are then made up, of which portions are finely ground for analysis. Table I gives details of the manurial treatment of the various plots in Broadbalk and Hoos Fields, in the soil of which the calcium carbonate has Table 1—Nature and Quantities per Acre of the Manures annually applied.* aa Ba || a $ g 3 6 | 86 eee] EE s|aelee| 4 | 4) 22] | 43 ae BS ge] sa | Se ze| ea | ge] os 3 2h 8 ag cess) |p Slew) EE) etsy | ee per ster S BOF aa Se \se | ea) & | 4/83) 32)/ a3 cr Sie ey Sa cet ON AV racy Macey taal fot coal rosy neta fer= tak | Broadbalk Field. tons. lbs. lbs lbs. lbs ewts lbs lbs lbs Sls dadaseneces 1844 | Unmanured |’ — — — _— = = = Bee Acca: 1844 14 — — — — — — = = Mgcounoonanee 1852 — — — — — 83 955 200 100 100 (qhatoonnenees 1852 — — 100 100 — 3°5 200 100 100 Wausae ecco ts 1852 — —_— 200 200 — 3°5 200 100 100 Sieichaeeecks 1852 — — 300 300 — 3°35 200 100 100 Ora rtaet sacri 1852 — 275+ — — — 3°5 200 100 100 ALO vrrehctne siciessers 1845 — — 200 200 — = = = — IE atone 1849 —_— — 200 200 — 3°5 [ices = = | i Hoos Field. NOG aire 1852 Unmanured | — — — — —_ — = — BOW asda 1852 — = — — — 3°5 200 100 100 ALEKS lanterns 1852 — 100 100 — = = — = AA ey cv 1852 —_ — 100 100 | — 3°5 200 100 100 Neyer oe 1868 — 275 — _— — = = = ANG ee sirsoeine 1868 — 275 — — — 3°5 200 100 100 IG e barra: 1852 — — — — 1000 -—— — = = (ee 1852 14 — — — — = — = = if * For certain minor variations in the amounts of manure applied see “ Memoranda of the Field and other Experiments at Rothamsted. Lawes Agricultural Trust, 1901.” ft 550 lbs. per annum up to 1884. been determined; Table II gives the average weight in pounds per acre of the fine dry soil in the layer 9 inches deep which is removed by the sampling tool for the different fields at each of the specified dates; Table III for Broadbalk, Table V for Hoos, and Table VII for Agdell and Little Hoos Fields give the percentages of calcium carbonate in the fine soils dried at 100° C. Mr. A. D. Hall and Dr. N. H. J. Miller. [Mar. 30, Certain difficulties are experienced in attempting to calculate from these figures the actual quantity of calcium carbonate per acre in the soil at different periods, owing to the impossibility of drawing samples that represent the same layer of soil on each occasion. Since changes of texture are set up by the different treatment of the plots the consolidation of the surface layer varies, so that the 9-inch slice includes more or less soil from time to time. On most plots the weights of the samples tend to get heavier, because the soil sets more closely together under the conditions of long-continued manuring Table Il—Weights of Fine Soil Dried at 100° C. per Acre. Broadbalk Field. Plots. Depth. 1856. 1865. 1881. 1893. 1904. lbs. lbs. lbs. lbs. lbs. OE ee Ne eh 1st 9 inches =! 2,200,000 | 2,400,000 | 2,400,000 | 2,400,000 Other plots ......... 1st 9 inches } 2,200,000 | 2,300,000 | 2,560,000 | 2,650,000 | 2,650,000 JA plots eee 2nd 9 inches | 2,590,000 | 2,590,000 | 2,590,000 | 2,590,000 | 2,590,000 All plots .....0..0.. 3rd 9 inches | 2,815,000 | 2,815,000 | 2,815,000 | 2,815,000 | 2,815,000 Hoos Field. | Plots. Depth. 1868. | 1882. 1904-5. lbs. lbs. lbs. PIR RASA Res, con to 1st 9 inches — 2,100,000 2,100,000. Other plots ......... 1st 9 inches | 2,400,000 | 2,400,000 | 2,400,000 JNU FTIOES soccascoanec 2nd 9 inches 2,721,000 2.721,000 2,721,000 INT FIO poncosseooce 3rd 9 inches | 2,891,000 | 2,891,000 | 2,891,000 Agdell Field. Plots. Depth. | 1867. | 1874. 1883-4. 1905. lbs. lbs. lbs. lbs. IND Faleta Gessoonedoce 1st 9 inches 2,140,000 | 2,400,000 | 2,500,000 | 2,500,000 INTIUFSGEG Soocopposnee 2nd 9 inches 2,450,000 | 2,450,000 | 2,450,900 | 2,450,000 Little Hoos Field. Plots. Depth. 1873. | 1904. lbs. lbs. All plots ............ 1st 9 inches 2,500,000 2,500,000 All plots ............ 2nd 9 inches 2,500,000 2,500,000 1905. | On the Retention of Bases by the Soil. id Table I1l—Broadbalk Wheat Soils—collected at various dates. Calcium Carbonate per cent. in Fine Soil Dried at 100° C. Keel : Sept., | Oct., | Oct., | Oct., | Sept., nk Manures. | 1856. | 1865. | 1881. | 1893. | 1904. 1st Depth (1—9 inches). | | (D> |} YG | je 1 p. ¢ 26 | Farmyard manure (14 toms) ...............eeeeee eee [== 4-20 7) 3°79 | 3-46 | 3°28 & || Wharrnensiny200ll Cosqauoosconnbeccectos beoeseees Soe ocseecnarers | 5°35 | 4:54 | 3:97 | 3°45 | 3°29 Smee Ullerminerals, se. deacssselseswiainrsctrica seacencnsswcets 5°65 | 4°96 | 3°75 | 3°34 | 2:94 6 a +200 lbs. ammonium salts ......... ; — | — 3°41 1°98 | 2°33 7 5D + 400 Ibs. mi | 8 noceandes | — | 3-82 | 3:19 | 2°36 | 2-25 8 x +600 lbs. OLR Sh eee & een en oEea elite 7o0 lot 76 9 i + 275* lbs. nitrate soda ............ | A249) 3-99 9 Se72)" || 3-36 10 | 400 lbs. ammonium salts only ................0000 | 5-41 | 4-10 | 3:31 | 2°76 | 2-47 11 55 0 and SHPO PEN. o| 4°36 | 3:14 | 2°76 _— i } | | 2nd Depth (10—18 inches). 2b | Farmyard manure (14 tons) ............00s0eseeees — | 0°277 | 0°310 | 0-422 | 0-237 & | havering cacosenobeqoos ono Goddoo Loe oso pEdade nna pedese — | 0°222 | 0-162 | 0:099 | 0:116 Ome eualemnin eras)” hectic ects cite otaku sisenedgercue senses — | 0:147 | 0-128 | 0:100 | 0-110 6 Fs +200 lbs. ammonium salts ......... } =e | = 0:110 | 0:114 | 0-132 fl 5 +400 Ibs. a, Decne esses — | 0:212 | 0:210 | 0:096 | 0-167 8 a +600 lbs. ih, BEAN Be p= — | 0-106 | 0-093 | 0-117 9 3 + 275* lbs. nitrate soda ............ | — | 0:3809 | 0:263 | 0:482 | 0°143 10 | 400 lbs. ammonium salts only ..............0..040. | — | 0-127 | 0:187 | 0:170 | 0°111 11 e Fe and superphosphate.... — | 0:119 | 0:179 | 0 107 | — 3rd Depth (19—27 inches). | | 26 | Farmyard manure (14 tons) ..............00 ee — | 0°181 | 0°121 | 0°130 | 0-095 fo Ge AML ney G8 206 Le soode der cao Bas ana mdocnbd BeleSeneeUnd aANOEOnES — |0°179 | 0:090 | 0-084 | 0-113 Ope pEiallimineralsy Woecnes as scelee sic sesce eneseie sencaseceee | — | 0-056 | 0:079 | 0:050 | 0-100 6 * +200 lbs. ammonium salts ......... } — — 0-058 | 0:073 | 0:112 7 ss +400 Ibs. te AER a, | — | 0-144] 0-115 | 0-075 | 0-136 8 és +600 lbs. he Pay aA eee — | 0-063 | 0-104 | 0-122 9 55 + 275* lbs. nitrate soda ............ — | 0:144 | 0:102 | 0:130 | 0-116 10 | 400 lbs. ammonium salts only ...................5 — | 0:073 | 0-070 | 0-070 | 0°105 11 5 3 and superphosphate...) — | 0-090 | 0-089 | 0°085 == * Double this amount applied from 1855 to 1884 inclusive. with saline manures and the gradual loss of organic matter. On the contrary, the soil of the plot receiving farmyard manure grows lighter through the great accumulation of organic matter. These changes of weight would be of little moment were the soil uniform, but as the calcium carbonate is almost wholly present in the surface soil down to about six or seven inches, small variations in the thickness of the slice taken by the tool cause a varying admixture of the poorer subsoil, and so may induce considerable change in the estimated 8 Mr. A. D. Hall and Dr. N. H. J. Miller. [Mar. 30, weights of calcium carbonate. It is impossible to eliminate wholly the errors thus introduced, especially in the case of Broadbalk, where the average weights of the 1865 samples are exceptionally low and where the subsoil shows also a good deal of variation. In Table IV two estimates have been drawn up for the amounts of calcium carbonate per acre. In the upper set of figures an attempt has been made Table [V.—Broadbalk Soils. Calcium Carbonate in Lbs. per Acre. Rate of loss per acre per annum. Plot. 1865. 1881. 1893. 1904. Whole | 1881—1904. period. Total of Ist and 2nd 9 Inches+a fraction of the 8rd 9 Inches, to bring them all to the same Total of 5,240,000 Lbs. of Soil per Acre. lbs. lbs. lbs. lbs. Ibs. Ibs. QD icceconiscs 100,400 99,300 94,300 85,100 392 618 Big eaee ap 110,800 105,900 94,000 90,200 528 683 55. seSaoaon0%04 118,100 99,400 91,100 80,800 809* 809 (\ Rostpogaconee — 90,200 55,400 65,200 1086 1086 Uaciiejecsinnaien' 93,900 87,200 65,000 64,000 767 1009 Geer tee = 75,500 48,300 49,700 1122 1122 Qe vawisctescnse 106,000 109,100 111,100 92,700 341 713 OY isonsgcoon06e 97,800 88,400 77,500 68,300 756 874 Tete 103,700 85,100 75,900 = 993 = First 9 Inches only, reckoned as weighing 2,500,000 Lbs. ZDesaeecsewace 105,000 94,700 86,500 82,000 590 552 33 ‘esondaosb0ar 113,500 99,200 86,200 82,200 800 739 SEA 124,000 93,700 83,500 73,500 878* 878 Gas: reds = 85,200 49,500 58,200 1174 1174 Ms saecad sys 95,500 79,750 59,000 56,200 1010 1024 Buss Boe Bs 71,000 43,200 44,000 1174 1174 Oe rainiejs cece 106,000 99,700 93,000 84,000 564 683 OM rear sicaee se 102,500 82,700 69,000 61,700 1045 913 a ea 109,000 78,500 69,000 a 1429 = * 1881—1904 only. to introduce a correction for the varying thickness of the slice by adding to the weights of calcium carbonate in the first and second depths such a pro- portion of the third depth as would ensure the comparison of an equal weicht of soil in all cases. The lower set of figures is based simply upon the per- centage of calcium carbonate in the upper layer of soil, assuming a general 1905.]| On the Retention of Bases by the Soil. 9 average weight of 2,500,000 lbs. for the fine dry earth in the top 9 inches of soil, thus leaving out of account both the changing weights of the upper slice and the contents of the subsoils. The two sets of figures lead to much the same comparative results, but the lower table is to be preferred as free from any speculative corrections, bearing in mind, however, that the rate of loss is probably over-estimated on most of the plots and under-estimated on the plot receiving farmyard manure. Fig. 1 shows these percentages of calcium carbonate in the upper soil plotted against the time. Fic. 1.—Percentage of Calcium Carbonate. Broadbalk Field. First 9 inches. A cursory examination of the figures and curves shows that if it may be assumed that the calcium carbonate was equally distributed over the whole field initially, then the greatest losses have occurred on the plots manured with ammonium salts, the loss increasing with each addition of ammonium salts; sodium nitrate, on the contrary, would seem to have exercised some preservative influence on the calcium carbonate, which is now at its maximum on the plot where that manure is used. Table VI gives the weights of calcium carbonate per acre in soils from the Hoos field as calculated from the percentages in Table V and from the average weights of soil in Table II, without any attempt at correction ; the results being also thrown into a graphic form in fig. 2. 10 Mr. A. D. Hall and Dr. N. H. J. Miller. [Mar. 30, Table V.—Hoos Field Barley Soils, collected at various dates. Calcium Carbonate per cent. in Fine Soil dried at 100° C. Plot. Manures. 1868. 1882. 1904-5. 1st Depth (1—9 inches). Per cent. Per cent. Per cent. nos aareacace Winimian Ure ercavavswoencaentoeseorenck netic moneee — 3°43 2 34 416) Gacwadoda Mall tnin erall’soiaccovanwece pounce scecsw asc teers 1:78 1°47 ©) 7/1 DA eserey sees 200 Ibs. ammonium salts ...........cc0eeeeeee ees —_ 3°26 2°53 ATKU ERE Cs Full minerals + 200 lbs. ammonium salts ... 1°34 0°87 0°28 AEN fasteners 275 Ibs. nitrate of soda ..........ccsescseeeeeseees — 3°30 2°59 AINienrenee Full minerals + 275 lbs. nitrate of soda ...... 1-30 1°32 0-48 ot eeerpoade 100 Tbs:\rape cakerecs..ssce-neeeevesesaseseaesen: — 1°58 0°89 ees 14 tons farmyard manure ...............06.e.000 — 1°92 1:'14 2nd Depth (10—18 inches). Gy eragsas Mnmanured encrcdcaasceeosctee sees See — 0-107 0 -202 MOwennstoac: Hall sminerallsh eanscccsostees cas nena eee ee meas 0°110 0-091 0:139 MUAY eter ce 200 Ibs. ammonium salts .............c.0seeeeeee — 0-116 0°119 AN Neer meat Full minerals + 200 lbs. ammonium salts ... 0-154 0-074 0-083 TENT aves 275 lbs. nitrate Of Soda ........500..esseescensn ess — 0°107 0-112 AN ceneesnen | Full minerals + 275 lbs. nitrate of soda ...... 0-064 0-073 0-112 MO osoens000 1000 Ibs. rape cake .........0.......0scccceseneee ens — 0-095 0 066 7T—2 14 tons farmyard manure ................2.60000+ = 0-275 0-147 3rd Depth (19—27 inches). MOU setaee LUhatre te hahvnys\s engnnnser seacododosen nbabcasedonoocabocsc6 — 0 :056 0-078 AO: sites Full minerals -asrcscannsosaerearaeinnirnen eee 0 :094 0-081 0-091 AAS estes 200 Ibs. ammonium salts ..............00e0 eee ees —_ 0-063 0 062 ATAGD Sexsratrcts Full minerals + 200 lbs. ammonium salts ... 0 096 0-090 0:073 TEN Ragodaane 275 lbs. nitrate of soda .............0ceeceeeee eee — 0-075 0-084 ANE een Full minerals + 275 lbs. nitrate of soda ...... 0 ‘063 0-061 0 °103 I) aeopese0 NOOO Ms srapel Cake cwawscreccis sdecadeeeseceeeeeeches — 0 068 0 067 72 w, 14 tons farmyard manure ............ce0eeeeee eee — 0-074: 0-103 Unfortunately the figures show at once that the initial chalking of the Hoos Field has been very irregular, the plots numbered 1 show more than twice as much calcium carbonate as the corresponding plots numbered 4. As also in this field, samples only exist for two dates, 1882 and 1904—5, instead of the four which are available for the Broadbalk Field, less weight can be laid on the results for the individual plots. It is, however, noticeable that it is one of the plots receiving ammonium salts every year which now contains the least calcium carbonate. 1905. | On the Retention of Bases by the Soil. Table VI.—Hoos Field Soils, Calewwm Carbonate in Lbs. per Acre. Rate of loss Plot. 1868. 1882, 1904-5. per acre per annum. Total for Three Depths (27 inches). Ibs. Ibs. lbs. lbs. I: coossocosvs00o0 == 86,800 63,900 1000 Aone) Seen 48,431 40,100 23,400 675 AY Goo podedooboeHd = 83,200 65,700 760 4, a0 cb8000 000000 39,100 25,500 9,900 790 TL a ae es 84,300 67,600 725 Csi dial 34,754 35,400 17,500 G5 16 ke eae pas 42,500 25,100 755 eRe scisinaycasiane — 49,900 30,900 830 First 9 inches only, reckoned as weighing 2,500,000 lbs. 110) -sonesostoanatos = 85,750 58,500 1185 AR eh Aa 44,500 36,750 17,750. 723 SG AR a syateslseisistasie = 81,500 63,250 793 Lyk scala 33,500 21,750 5,750 750 IGN! soosassooneaden = 82,500 64,750 772 Zi en ae 32,500 33,000 12,000 554 NOW as tjcesort = 39,500 22,250 750 (22 ae aa 48,000 28,500 848 80000 lbs, IN IA pa0) 60000 lbs. 1868 Fic. 2,—Calcium Carbonate, lbs. per acre. 1882 1905 Hoos Field. To 27 inches. 11 12 Mr. A. D. Hall and Dr. N. H. J. Miller. [Mar. 30, Table VII gives the percentages of calcium carbonate found in the soils of the Agdell and the Little Hoos Fields, which are not specifically manured. Table VII.—Calcium Carbonate per cent. in Fine Soil dried at 100° C. Agdell Field. | 1867. | 1874. | 1883-4. | 1905. | 1st Depth (1—9 inches). Unmanured ... 5 815 5 984 5 102 4, 522 Unmanured ... 5 912 6 °414 4,927 4-567 2nd Depth (10—18 inches). Unmanured ... 0 °304 0-514 0-137 0-185 Unmanured ... 0 220 0 “767 0°134 07138 Little Hoos Field. 1873. 1904. | 1st 9 inches... 3-974 2-704 2nd 9 inches... 0 °145 0:118 Lastly in Table VIII the rates of loss in all cases are brought together for comparison. In calculating these rates of loss the earliest’ samples (1856) from the Broadbalk Field have been rejected as not comparable with the Table VIII.—Rate of Loss of Calcium Carbonate from Soil, lbs. per Acre per Annum. Broadbalk| Hoos Agdell | Little Hoos, wheat. | barley. | rotation. various. Wmmannuredeereacpeeenesers cere Cereeeereere cree 800 1000 { eae 1046 Mineral manures only...............s.sccssssseeeeee 880 675 - +200 lbs. ammonium salts ... 1170 790 bs +400 Ibs. i eel ORO = bs +600 lbs. ‘ eel LO = + 275 lbs. sodium nitrate ...... 565 465 200 lbs. ammonium salts only ... — 760 400 lbs. ) 6 40 1045 _— 275 lbs. sodium nitrate only ... — 725 Farmyard manure ................+. 590 830 Ra pekcakenscsnc cus sormacte swe secession cwatacieptecteetecne — 755 1905. | On the Retention of Bases by the Soul. 13 others, the rate of loss is then based upon the difference between the earliest and latest samples from each plot. B. Loss of Caleiwm Carbonate on Unmanured Land. It will be convenient to consider the unmanured plots together, so as to get an idea of the initial loss of calcium carbonate to the soil when there are no disturbing influences introduced by the manure. All the unmanured soils show the same general characteristics—a comparatively large amount of calcium carbonate in the top 9 inches, varying from nearly 6 per cent. on Agdell Field down to little more than 2 per cent. on Hoos Field. The second depth, however, contains as a rule between 0:1 and 0:2 per cent., only in a few special cases does it rise above the latter figure; in the third depth the proportion is 0-1 per cent. or less. As the Rothamsted fields are situated on almost the highest levels of the chalk plateau, it is difficult to suggest any reason for the restriction of the calcium carbonate to the surface soil other than its artificial origin, and this conclusion is confirmed both by the irregu- larity in the amount found in different fields and by its absence where the land has not been under cultivation. Samples were taken from three places on the adjoining Harpenden Common, where the same class of land occurs at approximately the same level as the experimental fields, but which has never been in cultivation and now carries a growth of poor grass, gorse, and bracken. The results :— Soil of Harpenden Common, 1 to 9 inches deep... 0°210 p. ce. CaCOs. s . 10 to 18 r coo OG a show that initially the soil of the Rothamsted Estate must have been practi- cally devoid of calcium carbonate. The successive determinations of the amounts present in the soil of the cultivated fields give the following rates of loss per acre per annum for the unmanured land: Broadbalk 800 lbs.; Hoos, 1000 lbs.; Agdell, 922 and 938 lbs.; Little Hoos, 1046 lbs. or an average loss of over 944 lbs. per acre per annum due to percolating water only. These rates are all probably too high, because of the gradual exhaustion of the organic matter and the resultant consolidation of the soil which has affected the later samples. But even after making all corrections that are possible for this source of error, the rate of loss would still amount to 800 lbs. per acre. These estimates may be to some extent checked by certain determinations made in 1896—8 of the amount of calcium compounds present in the water percolating through the drain gauges at Rothamsted. These drain gauges are 14 Mr. A. D. Hall and Dr. N. H. J. Miller. [ Mar. 30, blocks of undisturbed soil of 1/1000 acre in area, isolated by impervious walls from the surrounding land and maintained without vegetation since 1870. Table IX shows the proportion of lime in the water during 20 months Table [X.—Lime in Drainage Water percolating through 60 inches of Soil, September, 1896, to April, 1898. Drainage through | CaO in drainage Rainfall. soil, 60 inches water deep. per million. : inches. inches. 1896. September............... 8-077 6 °362 60-1 Octobereerr grees 4,°132 2 992 56 °3 November............... 1 °387 0 833 46 -1 December ............... 4-416 3 811 49 °8 1897. JANUary .....eeeeeeeees 2-031 1 585 48 -9 February ............... 2-925 3 °264 47 °9 4-197 2 589 48 -2 1-913 0 °320 48 °4; 1-718 0 :047 53 °3 2 °734 0 865 541 0:467 0 -024 44) °5 3-238 0:105 61°0 2-440 0872 60 °5 ] 0-960 0-001 _— 1-048 0-110 50 *4 3 503 3 060 56 °6 1898. January ...........---- 0-795 0 821 53 ‘1 February ............... 1-098 0-047 30°35 Marci tte 1-060 0: 492 51 °5 AoA, saqnooceoons 00800008 1 443 0 -082 43 °2 Average ......... 2-479 1-414 53 °5 (September, 1896, to April, 1898); the average concentration is 53°5 parts of CaO per million for the gauge with 60 inches of soil, which on the average percolation of 13:8 inches would give an annual loss of calcium carbonate of 300 lbs. Two causes contribute to make this figure low. In the first place the soil of the gauges is not very rich in calcium carbonate, determinations made on samples taken in 1870 from the land immediately adjoining gave 3:06 per cent., while two small samples bored out from the actual gauges in 1905 gave 1:88 per cent. for the upper 9 inches. Secondly, the air contained in the soil of this plot must be comparatively deficient in the carbon dioxide 1905. | On the Retention of Bases by the Soul. 15 necessary to bring the chalk into solution; as a rule the soil gases get richer in carbon dioxide the greater the depth, but the soil of the gauge is cut off from the subsoil and open to the atmosphere at the 60-inch depth. The long absence of any crop or manure will have reduced to very small limits the amount both of organic matter decaying to carbon dioxide, and the organic sulphur compounds which by bacterial action become sulphuric acid and leave the soil as calcium sulphate. These causes will co-operate to lessen the removal of calcium carbonate from the soil in the gauge, and as a matter of fact the concentration of 53°5 parts of CaO per million observed in its drainage is only about half of the concentration of the water running from the tile-drains beneath the unmanured plot in Broadbalk, which according to Voelcker’s and Frankland’s analyses (Table X) amounted in the mean to about 99 parts per million. But assuming that this latter figure represents the average proportion of lime in the drainage water from the unmanured plot, and that the average annual percolation through the soil of this plot is 10 inches,* equal to that through the 60-inch gauge, the annual loss of calcium carbonate per acre should amount to 400 Ibs. for the unmanured plot instead of the 800 lbs. found by analysis of the soil of the plot. The number of analyses, however, upon which the former estimate is based, is too small for great accuracy. Table X.—Broadbalk Drainage Water. Mean of 10 analyses by Voelcker and Frankland. Parts per million. Total solid | Lime and uO LEAS. matter. magnesia. 2 Farmyard manure (14 tons) ..............5 367 -2 123 G) gravel 4! |) Whaoneb mb oyerel sco sdocessogoodvaGaud ade coonoonnoees 227 °8 99 5 Mineral shonllygueaesceseesesasecrccccsccnoaas 329 °8 132 6 » +200 ]bs. ammonium salts ...| 450°3 171 7 » +400 Ibs. . yl eee| B42 “4 207 8 » +600 Ibs. a Sed GBS 222 9 » +275 lbs. nitrate of soda ...... 405 °7 126 10 400 Ibs. ammonium salts alone ............ 441 ‘8 173 11 is 4 +superphosphate...| 490 °4 197 In a paper by Creydt, von Seelhorst, and Wilmst on the composition of the drainage waters from an ordinary field tile-drained at a depth of * Ten inches was the estimate formed by Lawes, Gilbert and Warington, ‘Journ. Roy. Agri. Soc.,’ 1882, vol. 48, p. 24; Warington, ‘Trans. Highland and Agri. Soc.,’ 1905, vol. 17, 5th series, p. 168, estimates the drainage as somewhat more than 8-2 inches, while a comparison between the concentration of chlorine in the water from the unmanured plot and from the 60-inch drain gauge would lead to an estimated drainage of 9 inches. + ‘Journ. der Landw.,’ 1901, p. 251. 16 Mr. A. D. Hall and Dr. N. H. J. Miller. [Mar. 30, 53 inches, the annual loss of lime is estimated at 630 kilogrammes per hectare, almost exactly equivalent to 1000 lbs. of calcium carbonate per acre. Unfortunately, the proportion of calcium carbonate in the soil is not given, but the agreement with our figures for the removal of calcium ecar- bonate is very satisfactory. | Another consequence of some interest follows from these determinations of the loss of calcium carbonate from the unmanured plots. The analyses already quoted of the drainage water from the unmanured plot of Broadbalk Field show about 100 parts of lime per million, equivalent to a loss of about 400 lbs. of calcium carbonate per acre in the surface soil instead of 800 Ibs. estimated from the analysis of the soil. But the deep-seated waters of the chalk contain on the average about 150 parts of lime per million, a deep well at Harpenden, for example, yielding 158 parts per million.* Assuming that the percolation through the unmanured plot on Broadbalk represents the average percolation over the chalk area, then the removal of chalk by solution would be in the ratio of the concentrations of the two waters in question, 2.c., the average annual denudation of the chalk by solution alone would amount on the one estimate to 600 lbs., or on the other to 1200 lbs. per acre per annum. As the specific gravity of the chalk is about 2-2, and it contains from 95 to 99 per cent. of calcium carbonate, this would mean a lowering of the surface by solution alone at the rate of either 1/11000 or 1/5500 of a foot per annum. These estimates depend upon the assumption that the percolation through this unmanured plot of arable land represents the percolation over the whole area of the chalk, whereas ordinary crops or even grass would cause increased transpiration and allow of less percolation. But on the contrary, the figure adopted for the concentration of the lime in the drainage water from the unmanured plot is based on only a small number of analyses and would appear to be too low. The concentration of the drainage waters would be increased by any use of organic manures, by heavy cropping or by permanent vegetation, all of which would increase the pro- duction of the carbon dioxide causing dissolution in the soil. For example the analyses of the water from the plot receiving farmyard manure every year show an average of 123 parts of lime per million instead of 99 from the unmanured plot. This increased proportion of lime in the water percolating through ordinary land may be set off to some extent against the lessened percolation due to crops, but on the whole the evidence is in favour of the lower rate, so that a denudation of about 1/10000 foot per annum is a more probable figure. The accuracy of the sampling is not sufficient to enable any conclusion to be * Warington, ‘Trans. Chem. Soc.,’ vol. 51, 1887, p. 543. 1905. | On the Retention of Bases by the Soil. 17 drawn as to whether the loss of calcium carbonate fluctuates with the rainfall and percolation during each period. But the magnitude of the annual loss is somewhat surprising; assuming it to be only 800 lbs. per acre, then the Broadbalk Field must have contained at least 70 tons per acre of chalk at the beginning of the nineteenth century, and still contains so much that it will not be exhausted by the end of the present century. As, also, the rate of loss will probably fall with each reduction in the quantity present, dissolution being proportional to the surface exposed, the period of complete exhaustion will be considerably postponed. Since much of the value of the land agricul- turally, both in its fertility and in its ease of working, depends on the presence of calcium carbonate, it is clear that for the last 100 years the agricultural community have in this respect been living upon the capital accumulated by their forefathers, and are taking no steps to replace the inevitable depletion of this capital. C. Effect of Manures upon the loss of Calcium Carbonate. Most of the manured plots under investigation receive a dressing of “mineral manures” in addition to the varying amounts and compounds of nitrogen. This mineral manure consists of 34 cwt. per acre of superphosphate containing 17 per cent. of soluble phosphoric acid (equivalent to 37 per cent. of calcium phosphate “made soluble ”), 200 lbs. of potassium sulphate, and 100 lbs. each of magnesium and sodium sulphates. Of these substances the superphosphate reacts immediately with the calcium carbonate of the soil, the sparingly-soluble di-calcium hydrogen phosphate being precipitated wherever the superphosphate solution comes in contact with a particle of chalk in the soil in accordance with the equation CaH,P203 + CaCO3 = Ca2H2P20s3+ CO2+ H20. To complete this reaction, the 34 cwt. of superphosphate would require about 47 lbs. of calcium carbonate, but so small an annual loss would hardly be perceptible in the analyses. The neutral sulphates of potassium, sodium, and magnesium should occasion no loss, for though they react with calcium carbonate, the resultant alkaline carbonate is retained by the soil and would be estimated as calcium carbonate by the method of analysis adopted, which is based upon the carbon-dioxide evolved on treating the soil with acid. The action of the plant also, discussed later in this paper, would probably result in the reconversion of the sodium and potassium carbonates into calcium carbonate. The action of the mineral manures, as seen in the analyses of the soil, has not occasioned sufficient loss of calcium carbonate to be apparent within the limits of accuracy VOL. LXXVII.—B. Cc 18 Mr. A. D. Hall and Dr. N. H. J. Miller. [ Mar. 30, of the determinations. In the Broadbalk Field, Plot 10 can be compared with Plot 7; both receive the same amount of ammonium salts, but Plot 7 receives the minerals in addition. The rate of loss is practically identical, 1010 lbs. per acre per annum on Plot 7 and 1045 lbs. on Plot 10. Unfortunately Plot 5, receiving minerals only without nitrogen, cannot be compared with the unmanured Plot 3, because both Plots 5 and 6 show an entirely exceptional rate of loss, not to be accounted for unless it be that one of the dells, from which the field was originally chalked, lies in these two plots, and may have caused a very irregular distribution of the chalk. But if the rate of loss on Plot 5 be calculated over the period 1881—1904 only, it amounts to 880 lbs. per acre against 800 lbs. per acre on the unmanured plot. In the Hoos Field the results are too irregular to bear much discussion, for the plots which receive minerals—4o, 4a, 4N—start with less than half the chalk contained by the corresponding 1o, 1a, and 1N. In the plots, however, which receive ammonium salts as a manure, the loss of calcium carbonate is much increased. A reaction of the type (NH4)2SO4 + CaCO; => (NH4)2CO3 + CaSO. takes place as soon as the ammonium salts are dissolved, the ammonium carbonate is adsorbed by the surface action of the humus and the finer clay particles of the soil until it is nitrified, while the calcium sulphate passes forthwith into the drainage water. When heavy rain follows the application of the ammonium salts to the Broadbalk wheatfield, only traces of ammonia find their way into the drains, whereas there is an immediate great increase in the calcium sulphate and chloride present in the drainage water. At Rothamsted the manure termed ammonium salts consists of an equal mixture of ammonium sulphate and chloride, and is applied to the various plots at the rate of 200, 400, and 600 lbs. per acre, quantities which would react with 161, 321, and 482 Ibs. respectively of calcium carbonate. Before, how- ever, the ammonium carbonate thus produced has been long in the soil a second molecule of calcium carbonate must be consumed to provide a base for the nitrous and nitric acids formed by its nitrification. This would make the loss caused by the application of 200 lbs. of ammonium salts up to 321 lbs. of calcium carbonate, half of which is caused by the initial reaction producing ammonium carbonate, and half by the nitrification of the latter. In order to determine the rate of loss in the field, it will be convenient, in the various cases where a comparison is possible, to subtract the average rate of loss on a plot without ammonium salts from the plots receiving ammonium salts, and divide by 2 or 3, as need be, to find the effect in each case of 200 lbs. of ammonium salts. 1905. | On the Retention of Bases by the Soul. ug Annual loss of calcium carbonate. Broadbalk (Plot 7—Plot 3)+2......... 105 lbs. per acre. i (Plot 10—Plot 3)+2......... 122 3 3 (Plot 8—Plot 3)+3......... 125 Ss Hoos (Plot 44—Plot 40)+ 1 ...... 115 rs Mean...... iiy/ 7" This mean value for the loss caused by 200 lbs. of ammonium salts is not very far from the 161 lbs. estimated above as necessary to convert them into ammonium carbonate, but neither the mean nor any of the individual analyses support the view that a second molecule of calcium carbonate is removed from the land by the nitrification of the ammonium carbonate. However, this nitrification must take place; indeed, there is every evidence that it takes place so rapidly and thoroughly in the Rothamsted soil that no ammonium salts are carried forward un-nitrified from one season to the next. The analyses of the drainage waters (Table X), while they show a progressive increase in the amount of lime for each addition of ammonium salts in the manure, do not permit any estimate to be formed of the rate of removal, so much is the extent of the percolation, as seen in the relative frequency with which the drains run, affected by the size of the crop, which becomes large on the heavily-manured plots. It remains, therefore, to be explained why the loss of the soil should be at the rate of one rather than of two molecules of calcium carbonate for every two molecules of combined ammonia applied in the manure. The plots receiving sodium nitrate in place of ammonium salts show not only no special loss of calcium carbonate due to the nitrogeneous manure, but a distinctly diminished loss as compared with the unmanured plot. On Broadbalk the nitrate plot loses at the rate of 564 lbs. per acre against 800 lbs. on the unmanured plot; on Hoos the loss is 465 lbs., against 679 lbs. on the plot receiving the same minerals but no nitrate. That the sodium nitrate exercises some conserving influence on the calcium carbonate in the soil is also apparent from a consideration of the analyses of the subsoil. On Broadbalk, where the subsoil from 10 to 18 inches contains in most cases 0:1 or 0:15 per cent. of calcium carbonate, the same layer beneath the plot receiving sodium nitrate shows 0°31, 0:26, 048, and 0:14 per cent. at the various dates, while the third depth of this plot is also richer than on the corresponding plots. Further evidence may be derived from the composition of the drainage waters (Table X); the water from the plot receiving mineral manures only Cc 2 20 Mr. A. D. Hall and Dr. N. H. J. Miller. [Mar. 30, contains 132 parts of lime per million, the plot receiving the same minerals and sodium nitrate gives a drainage water containing 126 parts of lime per million. Thus where the nitrate is used there is both a lower concentration of lime in the drainage water and a smaller total percolation, because of the much greater crop, and consequently increased transpiration cn this plot. The sodium nitrate then either saves the calcium carbonate of the soil from its normal loss or has some power to bring about the re-formation of calcium carbonate. With this fact must be correlated the non-disappearance from the other plots of the calcium carbonate required to form calcium nitrate with the nitrified ammonia base. Considering lastly the plots receiving farmyard manure, the Broadbalk Field shows a much lower rate of loss on the plot manured every year in this way than on the unmanured plot, 590 lbs. against 800 lbs. per acre per annum. The corresponding plots in the Hoos Field hardly confirm this view, since both the plot receiving farmyard manure and that receiving rape cake, the only other organic manure employed, appear to be losing calcium carbonate at much the same rate as all the other plots. However, as the Hoos Field results rest upon determinations made at two dates only instead of four as in the case of Broadbalk, it is much more probable that the result yielded by the latter is trustworthy. The subsoil of the plots receiving farm- yard manure also show amounts of calcium carbonate above the normal—0:28, 0°31, 0:42, and 0:24 in Broadbalk, and 0:28 and 0:15 in Hoos Field. The drain beneath the farmyard manure plot on the Broadbalk Field runs but rarely, because the humus derived from the long-continued organic manuring of this plot is capable of temporarily absorbing any ordinary rainfall and then passing it gradually down to the subsoil without causing the drain to run. But the few analyses that have been made of the water draining from this plot indicate a lower concentration in calcium compounds than would be expected from the large amount of carbon dioxide produced by the decay of recent organic matter, and also from the considerable annual addition of calcium compounds in the manure itself. The composition of farmyard manure is very variable, but the mean of a number of analyses gives 0°6 per cent. of CaO, or an annual application of 190 lbs. per acre, of which the greater part is in combination with organic acids. There is, therefore, an addition of calcium compounds in the manure more than equivalent to their greater concentration in the drainage water, and the net result is a gain of calcium carbonate to the soil as shown by the diminished rate of loss on this as compared with the unmanured plot. We thus obtain three lines of evidence that there is some agency saving or re-creating the calcium carbonate of the soil: (1) the loss of calcium carbonate 1905. | On the Retention of Bases by the Soul. PAL induced by the use of ammonium salts is less than half that required for the absorption and subsequent nitrification of the ammonia; (2) where sodium nitrate or (3) where farmyard manure is applied, the rate of loss of calcium carbonate is below that of unmanured land. Further evidence that there must be under normal conditions some action at work protecting or renewing the bases of the soil may be gathered trom the continued fertility of many soils containing but a trace of calcium carbonate. The following analyses may be quoted of soils that have fallen under the personal observation of one of us, soils which, despite their very low content of calcium carbonate, have continued to give crops under arable cultivation for a long period. Table XI.—Calcium Carbonate per cent. in various Soils. Pomanton: London j London | Gault| Weald | Bagshot | Thanet Woburn Experimental Farm, clay. clay. | clay. | clay. sand. beds. Stackyard Field. Barley. Tocalit Wanboro’,| Ashtead | Alder Sen Bisl Woodnes- 1876 1908. | 1908. 1903. y: Surrey. | Common. | Holt. pager aaa borough. ‘| Plot 2a.| Plot 3.| Rotation. p- ¢. p- ¢. 1b @ || je p- ¢. p. ¢. p. ¢. 2G: b @ p. ¢. Ist depth ...| 0:°065 0:002 | 0:04 | 0-037 | 0-008 0-018 0:087| 0-051 | 0-070] 0-089 2nd depth...| 0-084 nil 0:16 | 0-012 | 0-016 0-010 0 066 044. 042 | 0-071 Another striking case is afforded by the Stackyard Field on the farm of the Royal Agricultural Society at Woburn, which has been under experiment since 1876. Table XI also shows a series of determinations of calcium carbonate in the soil of this field taken at the beginning of the experiments and in 1902. The amount of calcium carbonate present is exceedingly small, barely determinable in fact, yet the plots continue to yield normal crops, except those which have been manured with ammonium salts. The latter in recent years have become almost sterile, showing an acid reaction to litmus paper and refusing to grow wheat or barley unless they first receive a dressing of lime.* Now in all these cases, however low the proportion of calcium carbonate may be, the action of the percolating water must remove some of it, and the recurring process of nitrification also demands a base. Yet the small quantity of base available does not disappear entirely so as to render the soil unfertile, unless some specially calcium carbonate consuming material, like the ammonium salts, is employed asa manure. The continued fertility of such * See J. A. Voelcker, ‘Journ. Roy. Agri. Soc.,’ 3rd Series, vol. 10, 1899, p. 585, and vol. 62, 1901, p. 272. 22 Mr. A. D. Hall and Dr. N. H. J. Miller. = [ Mar. 30, soils almost devoid of calcium carbonate has long been a problem, but it now seems probable that the calcium carbonate and other bases which are required for nitrification are in some way returned to the soil as bases, and that when a ready-formed nitrate like sodium nitrate is used as a manure there is an addition of available base to the soil or a corresponding diminu- tion in the amount of calcium carbonate removed by the drainage water. Furthermore these or other agencies conservative of calcium carbonate are sufficient to maintain the quantity in the soil at the level for comparatively healthy growth. Of the possible conservative actions, two will be now con- sidered and evidence be brought to show that (1) the normal growth of plants leaves behind a residue of base in the soil, (2) the decay of plant tissues results in the production of calcium carbonate. II.—Errect or PLANT GROWTH ON THE REACTION OF THE SOIL. The plant, it is well known, does not take up the salts of the soil water in the proportions in which they are present in the solution, but exercises a selective action in favour of substances necessary to the nutrition processes, such as potash and phosphoric acid. And if the composition of the ashes of the plant be taken into account, it is clear that the selective action is exercised not merely on the salts with which the root is in contact but on their acids and bases considered separately. For example, from a solution of calcium nitrate the plant would withdraw more nitric acid than its equivalent of lime and from a solution of potassium sulphate more potash than its equivalent of sulphuric acid. When a plant is burnt the ash is usually alkaline, because the organic acids and any nitrogen present as nitrate in the plant are all driven off, leaving the bases as carbonates. But when a balance is struck between the acids and bases in the ash and when the nitrogen present in the plant before burning is calculated as an acid, since it all entered the plant as nitrate, the acids are generally to be found in excess. Warington,* indeed, has already pointed out with reference to the published analyses, that plants must retain more acids than bases. It does not appear to have been noticed, however, that such a result, by leaving behind in the soil a corresponding excess of base, must have an appreciable effect upon the reaction of the soil, although Knop and other early investigators have observed that the solutions in which plants are grown as water cultures become alkaline after a time. It is hardly possible to decide whether the excess of base is left behind in the soil water * “ Agricultural Students’ Gazette,’ 1899, p. 133 ; see Lawes and Gilbert, ‘Journ. Roy. Agri. Soc.,’ vol. 55, 1894, p. 640. 1905. | On the Retention of Bases by the Soul. 23 or excreted by the root cells. The nutrient constituents pass through the cell wall of the root hairs by osmosis until the concentration is the same on either side ; such substances as are required by the plant are then withdrawn from action by the protoplasm, thus lowering the concentration on the inside and causing a fresh diffusion of that particular substance through the cell wall. Looking at the question from the standpoint of the ionic hypothesis, the soil water would be a highly ionised solution, in which any given ion will pass by osmosis into the root-hairs as long as the plant maintains the tension of that ion lower on the inside of the cell than in the solution outside. As the plant is always transforming the nitrogen, sulphur and phosphorous from the condition of inorganic acids in which they enter into neutral or even basic organic compounds, and since it also as a rule utilises more of these substances- than of the metallic bases, such acid ions will be withdrawn by the protoplasm in greater quantity, and so must pass through the cell wall from the external solution at a greater rate than the corresponding basic ions, the necessary equilibrium being maintained by the carbonic acid excreted by the root cells.* From this point of view, when the root is drawing nutriment from a solution of a neutral nitrate, the nitric acid ions would be travelling inwards to the protoplasm and the carbonic acid ions outwards, so that the carbonate of the base might be considered as forming outside the cell wall. If, on the other hand, the salt be considered to move into the cell undissociated and there to lose its nitric acid to the protoplasm, the base must then be supposed to diffuse out again as carbonate. The net result, however, is the same under either hypothesis, viz., that after the plant has been growing for some time in a neutral solution it will have taken up an excess of acid and left a corresponding excess of base, now combined as carbonate, in the solution representing the water of the soil. The following Table (XII) shows the composition of several crops as per- centages and again recalculated as equivalents of hydrogen, phosphoric acid being reckoned as tribasic, since the soil solution will be mainly derived from tribasic phosphates in the soil. It will be seen that there is a considerable excess of acid in the plant, from which it follows that an equivalent amount of base was left behind in the soil. This base is in most cases nearly equivalent to the nitrogen taken in as nitrate, and calculated as calcium carbonate will amount to between 100 and 300 lbs. of calcium carbonate per acre. In other words the normal growth of farm crops leaves behind from the salts in the soil used for its nutrition about as much base as would have been previously required for the nitrification of * Wor a discussion of the electrical disturbances such an interchange would involve, see Kohn, ‘ Landw. Versuchs.,’ vol. 52, 1899, p. 315. 24 Mr. A. D. Hall and Dr. N. H. J. Miller. — [ Mar. 30, Table XII.—Basic and Acid Constituents of various Crops. Agdell Agdell Agdell rotation. rotation. rotation. Panne Wheat. Barley. Swedes. ys Complete Complete Complete oe 1856-1873. manure. manure. manure. § courses. 8 courses. 3 courses. Dry matter per acre...............0.5 A749 4124 3899 4812 Nitrogen per cent. in dry matter... 0°88 1-00 2°41 1°55 Ash per cent. in dry matter......... 4°74, 4:00 6°59 7 +24 Percentage Composition of Ash. iHerricloxid eWeceesse-ceeseeseessinns ess 0°37 0 64 0-96 1°32 IRWIN®. Soooo0boocc00Ds090G0Ng20 cap noAOBAGOO 4-41 7 82 13 *42 8°27 IMaemesiay face op-cessensesennceaces: 2°98 4°16 2°70 3 42 JEOUERAO, — Scrodoooaadoscsoosangnce000b0c009 16°21 19 33 36 °33 35 °59 DOA mehr cissulee sasaiseems cose att telomeres 0°39 1°67 4-27 3 87 Phosphoric acid .........-..s.+e000 9-09 12 -88 9-38 7:96 Sulphuric acid ..................cc000 2°56 3°58 12 -36 5°76 (Cli erStA® soonsecoaococos nao gaac0Nbaq000000 1°46 3°15 4°18 14°83 SHINER, Sacnadcoosndsscoc0 dco 00000b990000000 60 °58 46°47 1 04 16°54 Constituents reduced to Equivalents of Hydrogen, and Lbs. per Acre. TaleraaV OPAC) Goaccosebsoo sob nocdon500860 0-03 0 04 0-09 0°17 Wik EM eewcwnetecss be ceswissaisitelsevinaelecslece 0°36 0-46 1°24 1-03 MB OMESIA ae .iiseseteneceiste sceeednccese | @©88 0°34 0°35 0-60 IPOtas henna sneeweccneenacts je sceties ce cies 0°78 0°68 a99 2°64 S Oda sessintns deesseueneroes Scuecmeasrs 0:03 0-09 0°36 0 -44, Basesstotaly Seocccrsnacceceecsecsiere 1°53 1°61 4°03 4°88 Phosphoric acid ...............-.-++» 0°87 0:90 1-02 1°18 Sulphuric acid .............0c..eseeeee 0:14 0-15 0-80 0°50 @hlorine ee sccceccersncoscictcteecestece 0:09 0°15 0-30 1-47 INTEV OREN aectiristun ceiesioseesuescemereenues 3:00 2°97 6°74 5°35 Sillcar- iva woescaanasacsOenaceeessessesent 4°53 2°55 0:09 1°91 Acids, total, excluding silica ... 4°10 4:17 8 86 8 50 MotaltotibAasest esc tecste eee 1°53 1°61 4:03 4°88 Excess of acid (as hydrogen) ... 2°57 2°56 4°83 3°62 Equivalent to calcium carbonate 129 1238 242 181 the nitrates which entered the plant, as measured by the nitrogen finally contained in it. Although the many ash-analyses which have been made of farm crops afford conclusive evidence of this restoration of base to the soil, it seemed desirable to submit the fact to experimental verification. In 1903 water cultures were made of barley and cabbage in normal solutions which were 1905. | On the Retention of Bases by the Soil. 25 contained in cylinders of Jena glass holding about 3 litres; it was found that after growth had gone on for some time the solutions had gained in alkalinity by an amount equivalent to the nitrogen taken up by the plant, less the alkalinity of the plant ash. One example of these trials will suffice. A cabbage plant made 7525 grammes dry matter containing 2°705 per cent. of nitrogen = 02035 gramme nitrogen, equivalent to 0:01454 gramme hydrogen. The ash of the plant had an alkalinity equivalent to 0701082 gramme hydrogen. The culture solution gained alkalinity = 0:00558 gramme hydrogen, which with the 001082 gramme alkalinity of the ash makes 0:0164 gramme of alkaline hydrogen found in ash and solution, to correspond with the 0°01454 gramme of acid hydrogen equivalent to the nitrogen in the plant. In 1904 six cultures were started with wheat in similar jars on March 3, and growth was continued without changing the solutions until June 11, when the grain was fully formed. Three grains were sown in each jar and growth was extremely vigorous, but was continued a little too long, for at the close of the experiment no nitrate remained in the solution. For purposes of analysis a mixture was made of Nos. 1 and 2, and of 3 and 4, the remaining two being kept in reserve. Table XII{ shows the results obtained. In the initial solution there was a trifling excess of acid over base owing to the presence of a little ferric chloride, of which the acid alone is brought into the account. At the end it will be seen that there is an apparent gain of nitrogen, due in the main to dust (the greater part being present in the sediment which formed in each jar) and the seed (which contained Table XIII—Wheat Grown in Water Culture. Composition of Plant and Solution after Growth. Original Solution. | | Quantities | Equivalent to | taken. hydrogen. | | grammes. gramme. Tee eta | 42190 0 -0898 IWEO) coseacoonosapoaeoeee 0 6018 0 -0301 (ORK) ceccaccsapenbenendeereee | 28000 0 -1000 Total bases | = 0 -2199 N (as nitrate)............| 1 -4000 0 -1000 P.O; (monobasic) ...... 2 0974 0 -0295 SiG eine A Be eee 1 -2080 0 -0302 WO pears tuscchs astaecttnsees 21658 0 -0608 Total acid......| _— 0 2205 26 Mr. A. D. Hall and Dr. N. H. J. Miller. [Mar. 30, TABLE XIl1—continued. After Growth. Nos. 1 and 2. Dry Matter produced = 93°7 Grammes. Quantities found. Equivalent to hydrogen. ‘ : ieee ; Sediment. | Solution. Plant. Sediment. | Solution. Plant. Total. LoS ae ed alles zee aoeea| Boe hl SS oe es sid gramme. | grammes. | grammes. || gramme. | gramme. | gramme. | gramme. LENO) Fath aone — 0 5014 3 6644 —_— 0 01065 0 :07796 0 ‘08861 MgO ...... (0 -2082*) 0 -2318 0-1618 (0 01041) 0 01159 0 -00809 0 -08009 CORK Opascsneon 0-152 1 -1592 1 ‘4371 0 :005438 0 :04140 0 :05140 0 :09823 0 -01584 0 -06864 0 °13745 0 21693 ING eto 0 °1249 0 0103 1 °3540 0 :00892 0 00074 0 :09670 0 ‘106386 1240 Pe eoesoer 0 °2942 0 °5631 1 246 0 -00414 0 :00793 0 :01755 0 :02962 SOs ede — 0 6448 0 -4008 — 0 01612 0 :01002 0 :02614 0) ieeaa goaacnd — 0 °8114 1 °3461 — 0 -02285 0 03792 0 06077 0-01306 | 0-04764 | 0-16219 | 0-22289 | Bases im excess...........+ +0 :00278 |+0 -01600 |=0 02474 |—0 °00596 After Growth. Nos. 3 and 4. Dry Matter produced = 77-6 Grammes. KG Oe. sees —_— 0 -6736 3 4819 — 0 01483 0 07410 0 08843 | MgO ...... (0 °1296*)| 0:°2413 0 :2809 (0 :00648)| 0 -01206 0 -01154 0 -03008 | CaO sav occks 0 °2240 1 +2222 1 °3414 0 :00800 0 04365 0 :04790 0 -09955 0 01448 0 :07004 0 183854 0 -21806 WAN Wiiscesmsseee 0 :0972 0 :0080 1 4088 0 :00694 0 00055 0 10027 0 °10776 | 1B Oa Baotop 0 °345 0 °5765 1:170 0 :00486 0 -00812 0 -01648 0 :02946 SOnnn — 0 ‘6610 - 0 °3959 — 0 -01652 0 -00990 0 :02642 CCl lbeeseen coos —_— 0 8595 1 °2212 — 0 -02421 0 08440 0 ‘05861 | 0 :01180 0 :04940 0 °16105 0 22225 | Bases in excess......-..... +0 00268 | +0 02064 |—0:02751 |—0-00419 | * Estimated figure—these two determinations were lost. 00036 gramme nitrogen, equivalent to 0:00026 gramme hydrogen), and partly to a trace of ammonia in the large quantities of distilled water evaporated during growth. This excess of nitrogen accounts for the slight excess of acid over base in the final result when all the nitrogen is reckoned as acid. It will be seen that the solution and sediment (representing the soil) gained in the one case base equivalent to 0:01878 gramme of hydrogen and in the other to 0:02332 gramme of hydrogen, quantities which would probably have been increased had the growth not been continued until all the nitrate was exhausted. It has been often supposed that plants excrete some organic acid from the 1905. | On the Retention of Bases by the Soil. 27 root which aids in bringing about the solution of nutrient materials in the soil,* but no evidence could be found that anything except carbon dioxide passed from the roots into the culture solution. After growth a considerable bulk of the culture solution was evaporated to dryness and gently heated, very slight signs of charring were observed, no more than could be attributed to the dust, and the residue showed the same alkalinity before and after ignition, as would not be the case had any organic acid excreted from the root been present. On one occasion freshly precipitated ferric hydrate was introduced into the solution as a source of iron; although this was distributed all over the young growing roots so that it could not be shaken off, the plants began to suffer from lack of iron, and continued to do so until a trace of ferric chloride was introduced. Despite the well known acidity of the root- sap there seems no evidence that in normal cases it ever passes outside the cell wall, as long as the roots are unbroken. These experiments then afford experimental justification for looking to the growth of the plant as an explanation of some of the difficulties raised by the rate of loss of calcium carbonate on the different plots. The analyses of crops already quoted serve to show that the return of base to the soil may be large, quite sufficient to make up for the calcium carbonate required each year for nitrification. Hence soils which start with very small proportions of calcium carbonate may yet preserve their healthy condition and permit of nitrification, the losses caused by which are thus repaired. Again it becomes intelligible that the use of ammonium salts as a manure only occasions the loss of one molecule of calcium carbonate for each two molecules of ammonia, since the second molecule required for nitrification will be more or less restored during the growth of the plant. It has already been shown that the actual loss of calcium carbonate to the soil caused by the use of 200 lbs. of ammonium salts approximates to 161 lbs., and not the 322 lbs. which would be required if the calcium nitrate produced by nitrifica- tion were wholly removed from the soil. Further, when nitrate of soda is used as a manure, from the neutral sclution in the soil of calcium or sodium nitrate an excess of acid will be taken by the crop, leaving the soil richer in base. Hence the conservative action of sodium nitrate on the calcium carbonate of the soil that is visible in the analyses of both Broadbalk and Hoos Fields. It is possible to calculate the amount of base restored to the plots receiving nitrate of soda on the assumption that they possess the same average composition as the wheat and barley in Table XII, and that the amount of base returned * See Czapek, ‘ Pringsheim’s Jahr. f. wiss. Botanik,’ vol. 29, 1896, p. 321; Kossowitsch, ‘Ann, de la Science Agronomique,’ 2nd Series, vol. 1, 1903, p. 220. 28 Mr. A. D. Hall and Dr. N. H. J. Miller. [Mar. 30, will be in proportion to the size of the crop. In this way the following results are obtained :— Table XIV. Broadbalk Field. Hoos Field. Plot 8. | Plot 9. | Plot 40.} Plot 4x. lbs. Ibs. lbs. lbs. Total produce (grain and straw) ........ ...seescecsecenceweee 1936 6133 2343 5524 Bases restored to the soil as calcium carbonate, cal- ) culated from total produce ..........-.--.cecsccssecereen senses 53 167 73 ayal | | Mean rate of loss of calcium carbonate from the soil ... 800 564 675 465 | ——— | Total annual consumption of calcium carbonate ...| 853 731 748 636 The agreement between the figures in the last line is not very close, but indicates that the restoration of base to the soil, as calculated from the increase of crop on the plots receiving nitrate of soda, is approximately equivalent to the lower rate of loss of calcium carbonate found on analysis of the soil of these plots. The results, as a whole, go to show that the action of plants, in leaving behind a basic residue from the neutral salts in the soil upon which they feed, is a very essential feature in the chemistry of the soil, explaining, amongst other things, the maintenance of healthy conditions on the many soils poor in calcium carbonate. It also serves to explain one or two other points which have been observed in connection with the use of sodium nitrate as a manure. It has long been noticed that the continued use of sodium nitrate is very destructive to the texture of a clay soil, intensifying all the clay properties, rendering the soil persistently unworkable when wet, and forming hard and intractable clods when dry. The ultimate cause of such an effect is the “deflocculation” of the fine particles composing the soil; they are no longer bound together in loose aggregates, but are separated so as to give the soil its most finely grained character. Such deflocculation of the soil can be brought about by a trace of any soluble alkali, just as the opposite state of flocculation is induced by a slightly acid reaction. The Rothamsted soils continuously manured with sodium nitrate show marked signs of defloccu- lation, the drainage water from the nitrated plots in the Broadbalk Field is always more turbid than that from the other plots, and as one of us has shown,* there results in time on the nitrated plots a perceptible washing * Hall, ‘Trans. Chem. Soc.,’ 1904, vol. 85, p. 964. 1905. | On the Retention of Bases by the Soul. 29 down of the finest particles set free by the deflocculation into the subsoil or the drains. The bad texture of the soil following on the use of sodium nitrate is particularly to be seen on the mangel field, where it reaches its maximum on the plots receiving sodium nitrate and other neutral alkali salts like potassium sulphate and sodium chloride: it has been repeatedly observed to be at its worst in the winter and spring after a large crop has been grown on the sodium nitrate plots. As the soii of this field contains but little calcium carbonate, some of the base left behind in the soil by the growth of the crop would consist of bicarbonate of sodium or potassium, especially where the other alkali salts are applied in the manure, and there would be quite enough free alkaline carbonate thus formed to cause a thorough deflocculation of the soil. This explanation would agree with the observed fact that the defloccu- lation is much diminished where superphosphate only, an acid manure, is used in conjunction with the sodium nitrate. IT.—Errect or OrGANIC MANURES ON THE REACTION OF THE SOIL. Although the evidence is not so trustworthy as in the case of sodium nitrate, yet the use of farmyard manure and of rape cake seems also to result in a diminished rate of loss of calcium carbonate to the soil. Some of this may be due to the lessened percolation consequent on the greater water- retaining power of the soil enriched in humus, but another cause may be sought in the bacterial decomposition of calcium salts in the organic débris. Farmyard manure contains various calcium salts derived from the vegetable matter out of which it has been formed, sometimes in their original form, but partly broken down into the undefined carbon compounds known as “humates.” Calcium humate, Wollny* has already shown, can be converted into calcium carbonate by bacteria present in the soil, while the following experiments show that the commonest of all calcium salts in the plant, the widely distributed calcium oxalate, is readily fermented to carbonate. 100 c.c. of a nutrient solution containing— Ammonium sulphate ............... 0: 2 gramme OCMC MI OTIC Mewes swectee acc of aie esac 0:2 os Potassium hydrogen phosphate... 0-1 y Magnesium sulphate ............... 00555 Hermous SUP Mates cess sceacsee eet 004 , were placed in an Erlenmeyer flask plugged with cotton-wool in the usual way; to this 1 gramme of calcium oxalate was added, together with, in some cases, a small quantity of other organic nutrient, and the flask and its * ‘Zersetzung der organischen Stoffe,’ 1897, p. 217. 30 Mr. A. D. Halland Dr. N. HJ: Miller. [Marae contents were sterilised; when cool they were seeded with 0:2 gramme of partly dried surface soil recently drawn from Plot 2 on Broadbalk Field, and the flasks placed in a dark cupboard at the ordinary laboratory temperature. The results are summarised in Table XV, and serve to show that the soil XV.—Bacterial Decomposition of Calcium Oxalate. l Refer: | CaCo3 found | ence | Added to nutrient | Duration, | from |Reaction| State of nitrogen compounds No solution at starting. days. 1 gramme | at end. at end. : CaC,0,. | ¥ | | 46 | No soil added ......... 79 — = 50 | No calcium oxalate ... 79 0-015 Neutral 61 5 % baa 176 | trace Es Strong nitrite, slight nitrate. Ammonia. 44 | Neutral ...............065 "9 0-263 “3 | Nitrate, no nitrite nor am- monia. 62 ios) ude Sates eeerene 176 OGG i 5 Nitrate, slight nitrite, no | ammonia. 43 | Slightly alkaline ...... 73 0 428 4 Strong nitrite, nitrate. 64 3 5 oe 176 0-274 | 4, Slight nitrite, strong nitrate, no ammonia. 42 +0:2 glucose, neutral 79 0-105 ¥ No nitrite nor nitrate. 58 | +02 ,, 5 78 N)05F} |) gg No nitrite nor nitrate. Am- monia. Glucose gone. 52 | +02 ,, alkaline 73 0-184 | Ss Little nitrite, no nitrate, no } ? glucose nor ammonia. 48 +0°2 peptone ......... 80 0°183 59 Nitrite, strong nitrate, no am- | monia nor organic matter. SOR Oe es) gi stasiiean 84 0-184 | “ Strong nitrite and nitrate, some ammonia, no pep- | tone. 47 | +0:2 calcium humate 82 Og || 5 | Nitrite and nitrate, no am- | _ monia. 60 | +02 F ; 176 O1G5 | Slight nitrite, strong nitrate ? > § x g 5) no ammonia. 63 | +02 «4, 5 176 O11440" Pe—,, Slight nitrite, strong nitrate, no ammonia. contains one or more organisms which are very effective in converting calcium oxalate into carbonate. The mechanism of the reaction is being further studied; in the present connection the experiments are sufficient to show the existence of other agencies of a bacterial nature engaged in restoring calcium carbonate to the soil. The destruction of nitrates by bacterial action, with the evolution of the nitrogen as gas, the change commonly known as “ denitrification,” is always attended by the production of a carbonate of the base with which the nitric acid was combined, but as any calcium carbonate formed in this way would only replace the calcium carbonate consumed in the previous nitrification there would be neither gain nor loss to the soil. As also denitrification is most likely to take place in the lower subsoil where the oxygen of the soil 1905. | On the Retention of Bases by the Soil. 31 gases has been exhausted, any calcium carbonate re-formed in this way would not appear in the analyses set out, which only extend to the depth. of 27 inches. It is, however, clear that manuring with organic manures, the growth of clover and other leguminous plants which leave behind a considerable residue of roots and stubble particularly rich in calcium oxalate, the débris of plant tissues which accumulates in the soil of grass land, all go to maintain the stock of calcium carbonate, which in its turn is being as constantly drawn upon for nitrification and for the neutralisation of the other acids produced during the bacterial decay of the carbon compounds the soil receives. Doubtless in all soils containing only a minimal amount of calcium carbonate under natural conditions these various actions have reached an equilibrium, since the increase of any one only tends to bring into play the factor which limits it (the rate of nitrification, for example, will be slowed down as the available base in the soil becomes scarce), but also accelerates the operation of some action in the opposite sense; even the one irrevocable loss by drainage and removal of crop will probably be balanced by the calcium salts coming into solution through the continued weathering of the soil particles. In the main, however, the original stock of calcium carbonate in the soil circulates continually between plant and soil without suffering appreciable loss. It is only under particular conditions, such as the use of ammoniacal manures, or the setting up of anaérobic conditions through lack of drainage, thus allowing the formation of organic decay acids but not their final oxidation to carbonates, that the soil will develop an acid reaction and become infertile. Summary. The chief points brought out in the course of the investigation are as follows :— (1) Arable soils which contain upwards of 1 per cent. of calcium carbonate are subject to a normal loss of that constituent in the drainage water amounting to about 800 lbs. to 1000 Ibs. per acre per annum. (2) The loss is increased by the use of ammoniacal manures by an amount equivalent to the combined acid of the manure. The loss is diminished by the use of sodium nitrate or organic débris like farmyard manure. (3) The growth of plants normally returns to the soil a large proportion of the bases in the neutral salts which the soil provides for the nutrition of plants. (4) The calcium oxalate and other organic salts of calcium present in 32 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, plant residues are converted by bacterial action in the soil into calcium carbonate. (5) The return of base by the growth of plants and the production of calcium carbonate by the decay of plant residues are sufficient to maintain soils neutral which are poor in calcium carbonate, and to replace the bases which have been consumed in nitrification and similar changes. On the Origin and Life History of the Interstitial Cells of the Ovary wn the Rabbit. By JANET H. LANE-CLAYPON. (From the Physiological Laboratory, University College. Communicated by Professor E. H. Starling, F.R.S. Received June 16, 1905.) [PLATE 1.] The majority of the investigators of the subject consider that the cells of the germinal epithelium arise by differentiation of the peritoneum, and become embedded in the subjacent mesoblast, there being probably a dual process, namely, the downgrowth of the cells themselves and a simultaneous upgrowth of the subjacent mesoblast. The fate of the cells thus embedded has given rise to much discussion. All observers agree in stating that they give rise to the ova, and most observers consider that they give rise also to the follicle cells; but de Foulis (8),* Schrén (16), and Wendeler (20), believe these cells to be derived from the connective tissue. Pfliiger (15) and Waldeyer (18), although differing in regard to the develop- ment of the ovary, both consider that the germinal cells give rise to the cells of the follicular epithelium, there being most probably a previous division of the original cells. Nagel (14) also agrees that the follicle cells are derived from the germinal epithelium. Balfour (4) believed that some of the cells of the egg-clusters became ova by differentiation, and he described besides a number of small cells, of which some formed the follicular epithelium and the others probably either served as foodstuff for the rest, or eventually themselves formed ova or follicle cells. Bih' r(7) c seribes the formation of the follicular epithelium by the * Th se numbers refer to the entries in the bibliography at end. a 1905.| the Interstitial Cells of the Ovary in the Rabbit. 33 streaming inwards from the periphery of some of the cells of the germinal epithelium. The changes connected with ovogenesis have been very fully described by v. Winiwarter (19). They may be briefly summarised as follows. The germinal cells of the second invagination are rather small and show a nucleus with some lumps of chromatin, being also rather granular. These he calls protobroque cells of the a variety. These divide, giving rise to other protobroque cells a, and also to a b variety. These last divide again giving rise to more cells of the 0 variety and to a new form of cell, deuto- broque. The last are larger, and the nucleus more transparent. The deuto- broque cells give rise to the ova by nuclear differentiation by means of the following stages. 1. The chromatin breaks up into fine filaments, which are distributed over the whole nuclear area; this is the /eptotenic stage. 2. The filaments become gradually massed together until they show as a compact lump at one side of the nuclear area. This transformation is the synaptenic stage, which is succeeded by 3, the pachytenic. Here the filaments become again spread out, but they are much coarser than in the previous stages The 4th stage, or diplotenic, is so called on account of the tendency of the chromatin strands to lie in pairs. In the final or dictyate condition the chromatin is distributed in a reticulum over the greater part of the nuclear area Balfour describes protoplasmic masses of young ova where the cells appear fused, and he suggests that one of these ova may grow at the expense of the rest. Wan Beneden (5) describes multinucleated masses in the ovary of the adult bat, which he suggests may give rise to an ovum and its follicular epithelium. The formation of follicles, which proceeds rapidly, gives rise in the ovary to two zones, an external or parenchymatous zone in which the follicles lie, and an interstitial vascular zone; these have been described with some modi- fications by various workers and for different animals. (His(11), Waldeyer, Born (6), Macleod (13), Van Beneden.) The question of the post-natal formation of primordial ova has been the subject of many isolated observations. Pfliiger believed he had evidence of the return of the ovary to the tubular formation at the rutting season, the object of the return being the formation of fresh ova. Waldeyer believed that all ova were formed in the young animal, and for this reason called all ova “ primordial ova.” Schrén noticed an increase in the number of clear cells, presumably ova, near the periphery in cats and rabbits at the rutting season, and in women at the menstrual periods. Koster (12) describes prolongations of epithelium with formation of fresh ova and follicles in the ovaries of evera’ -ecently VOL. LXXVIL—B. ; D 34 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, pregnant women; Wagener(17) records the thickening of the germinal epithelium near the attachment of the Fallopian tube in the pregnant bitch, a condition which he thinks denotes the formation of ova. Amann (1) describes the presence of follicles in the ovaries of a woman of 65, where there was incipient cystadenoma, the follicles being in all stages of formation by means of invagination of the germinal epithelium. The interest of this observation lies in the age of the woman and in the apparent formation of fresh ova consequent on the stimulus caused by the incipient eystadenoma. V. Winiwarter was not able to trace any of the stages of ovogenesis in any of the adult ovaries he examined, and considers this a necessary condition for the formation of ova. As far as the literature goes, we may consider the ovary to be formed by the embedding in the underlying mesoblast of the cells of the germinal epithelium, the embedding being brought about by a process of ingrowth of the cells and of upgrowth of the mesoblast. The cells thus embedded are oogonia, which give rise to ova by division, as also to the follicle cells, the future ova undergoing considerable nuclear trans- formation before reaching the condition of the fully-formed primordial ovum. The post-natal formation of primordial ova has been recorded in certain cases, but there is not much evidence either in favour of or against it. Olject of the Investigations.—Certain features which I observed in the interstitial cells of the ovaries of rabbits at alate period of pregnancy led me to study the origin of these cells. This question would appear to have been neglected by previous workers on the ovary. The formation of an internal ovarian secretion (cf. Andrews (3)), which by analogy with the interstitial gland of the testis might be presumed to be derived from the interstitial cells (¢f. Ancel and Bouin (2)), gives considerable interest to their origin. This was studied by examining (1) the ovaries of rabbits from the twentieth day embryo up to those of the young rabbit about three weeks after birth ; (2) the ovaries of pregnant rabbits at all stages. Methods.—It is not easy to find a really good fixing agent for ovaries, especially adult ovaries. Hermann’s, Flemming’s (strong formula), Pod- wyssoski’s and Altmann’s fluids, were all used. The last was found satis- factory for cytoplasm, but the sections obtained with the other fixatives were not good. The tissue was osmicated outside and insufficiently fixed inside. Finally,* Gilson’s fluid was used exclusively for all nuclear figures, and a * Gilson’s fluid = abs. ale., 1 part ; glacial acetic, 1 part ; chloroform, 1 part ; the whole saturated with sublimate. 1905.] the Interstitial Cells of the Ovary in the Rabbit. 35 mixture of sublimate (saturated) 4 pints, formol 1 pint, and 1 per cent. glacial acetic for other purposes. The sections fixed in Gilson’s fluid were stained with iron hematoxylin, or heemalum ; those fixed in the other solution stain well either with iron hema- toxylin, hemalum and eosin, or toluidine blue and eosin. Changes in the Cells of the Germinal Epithelium in the Immature Rabbit. The origin of the germinal epithelium from the peritoneum by a process of differentiation has been so fully shown by several observers, that it will not be necessary to deal any further with the origin of the germinal cells. Also it has been shown that these cells become embedded in the underlying mesoblast ; this state of affairs is seen in an embryo of the twentieth day. The ovary is by this time a definite organ; it is intensely vascular, showing large blood spaces, especially in the parts lying immediately round the mesoblastic core. At this period the main mass of the germinal cells is situated peripherally, only a few isolated ones having penetrated into the core, which last is sending processes of connective tissue in between groups of germinal cells. Of these there are present a large number of protobroque and a few deutobroque ; also a certain number of mitotic figures, but these are not numerous. (See Plate 1, fig. 1.) From this time onward until after birth the changes in the ovary, as seen under the low power of the microscope, are not striking; there are more deutobroque cells, characterised by their transparent appearance, and there is an increase in the number of mitotic figures. Studied under the high power of the microscope some of the deutobroque cells are seen to have entered upon the early stages of ovogenesis, and to have reached the leptotenic stage. There are large numbers of round cells showing a nuclear structure differing from either the protobroque or the deutobroque cells. The mitoses are chiefly found near the periphery, and the greater number of them seem to be taking place in the large cells. I do not altogether agree with v. Winiwarter on tke question of mitosis in these cells. In the first place there appears to be very little distinction between the varieties of protobroque cell, a and 0, and I shalJl not dwell upon it. The mitosis in the protobroque cells does not appear to be sufficient to account for the large number of deutobroque cells which are formed, and my observations are to the effect that by far the greater number of mitoses are taking place in the deutobroque cells themselves. Each class of cell divides, the protobroque less copiously than the deutobroque, giving rise to two cells of their own variety. There can be no doubt that the deutobroque cells are modified D 2 36 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, germinal cells, but I hope to show that the process is one of differentiation, and not of division. The protobroque cell is the type of the original germinal cell ; it is small, generally oblong or oval, and contains a Jarge nucleus. The nucleus shows a number of chromatin masses of varying sizes, and the whole nuclear area gives a general impression of granulation (represented by shading in the figures (Plate 1, fig. 2 (a)). There are no chromatin filaments. The deutobroque cell is very much larger, and the nucleus has for the most part a strikingly transparent aspect, the granular appearance noticed in the protobroque cell being confined to the periphery of the nuclear area. The chromatin is quite differently arranged; there are one or two irregular chromatin masses, and strands showing nodular enlargements where they intersect. See fig. 2 (/). ‘ Sections of a young ovary very soon before or after birth show a large number of cells whose nuclei exhibit every phase of transition between these two varieties. These changes in the nucleus may be classed broadly into three divisions :— 1. The chromatin masses become fewer and larger. 2. There is considerable formation of chromatin strands. 3. The granular appearance gradually passes away from the centre of the nucleus towards the periphery, leaving the centre clear. Some of these changes are shown in fig. 2 (0), (c), (d), (e). In the first stage the whole cell becomes rounder, as also the nucleus, and the chromatin has begun to aggregate, and there are traces of strands passing away from the masses. These features increase in intensity until there are only a few chromatin masses, but the strands are passing between them and intersect in parts. The granular appearance has: begun to leave the centre, which is clear. A further process on these lines brings the cell into the typical deutobroque condition. It would therefore seem that the change from the protobroque type is accomplished by means of transformations in the nuclear area, accompanied by a growth in size of the cell, and it is unnecessary to suppose, under these circumstances, that mitosis is also a method of formation. The protobroque and deutobroque cells are therefore all oogonia, either potential or actual, the transition from the one class to the other being probably accomplished by processes of nuclear differentiation. At this period in the life of the ovary there is no appearance which could be characterised either as egg-tubules or egg-clusters; there are large collections of epithelial cells bounded centrally by the mesoblast, which presents the appearance of connective tissue. This tissue penetrates but slightly into the region lying peripherally to the main central core, but careful 1905.] the Interstitial Cells of the Ovary in the Rabbit. 37 inspection shows that there are a few fine processes pushing their way outwards and more or less enclosing large numbers of germinal cells. The latter are of all shapes and sizes, from the typical protobroque to the deuto- broque type. By the third day after birth the general configuration of the ovary has changed very considerably. There is still the central mesoblastic core, but the germinal cells have become more marked off than in the embryo, presenting the appearance of a definite zone of germinal epithelial cells. The cells are arranged, especially in the more central parts of the zone, in the form of solid rods or clusters, several cells thick, which press their rounded ends into the central mesoblast. Some of them might fairly easily be mistaken for tubules without a lumen, and there can I think be little doubt that these are what Pfliiger took for tubules. Around the periphery the tubular formation is not so marked, the cells lying in irregular aggregations (fig. 3). As the mesoblast is centrally situated those parts of the germinal zone lying towards the centre get divided up earlier than the more peripheral parts, which retain the formation of an earlier stage. This lagging behind, as it were, of the periphery is quite characteristic of all the changes taking place in the young ovary; it applies to the formation of tubules and clusters, to the processes of ovogenesis, as pointed out by v. Winiwarter, as well as to the formation of interstitia cells, which will be dwelt upon later on. From this time onwards up to about the twelfth day after birth the changes in the general contiguration of the ovary are brought about by an amplification of the processes already described, namely, continued upgrowths of connective tissue, cutting off the tubules and clusters. The connective tissue likewise presses into the larger collections of germinal cells, thus cutting them off and dividing them again into smaller portions, so that as time goes on the clusters near the central parts consist of less cells, but are present in much greater number, while those parts more peripherally situated are in a somewhat earlier stage, the cluster formation being still fairly evident close to the periphery as late as the sixteenth day.* Before proceeding to the changes in the egg-clusters about the fifteenth * Too much stress, however, should not be laid upon the exact date of the young ovary in relation to its structural aspect. There seems to be an appreciable difference in the extent to which the ovary is developed in different animals about this age. v. Winiwarter does not describe any ovary between, the tenth and the eighteenth day, because he does not consider the changes to be sufficiently striking to call for any description. Of two litters of rabbits I found slight differences in the ovaries of the same date, the sixteenth day, the changes being rather more advanced in one than in the other, and both were almost as advanced as v. Winiwarter’s figure of the eighteenth day. The differences are probably determined by the varying nutrition of the animal, as also possibly by the kind of rabbit, some being far more advanced in outward aspect at this age than others. 38 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, day, the transformations which have been taking place in the deutobroque cells must be briefly touched upon, but they have been so fully described by v. Winiwarter that a lengthy exposition is quite unnecessary. I shall adopt his nomenclature throughout. It has already been stated that almost immediately after birth changes begin to take place in the deutobroque cells, which enter upon the leptotenic phase, the transformations beginning centrally. This is succeeded by the synaptenic, and by the third day there are already a very great number of this variety. The leptotenic phase is characterised by the absence of visible nucleolus, and by the spreading out of the chrcmatin in the form of fine filaments over the whole nuclear area. This stage is evidently only a further step in the differentiation which has already taken place. It has been shown that the change from the protobroque to the deutobroque type is accomplished by the chromatin masses becoming gradually broken up into strands. In the leptotenic phase the process is merely carried further. Whereas in the deutobroque stage there are still one or two chromatin masses which have not become broken up into strands, in the leptotenic this is not the case, the whole chromatin being present in the form of filaments. The transition stage can be seen in an ovary of a few days after birth (fig. 6(1)). These filaments become gradually aggregated, passing to the synaptenic state. The leptotenic condition is a very fugitive one, whereas the synaptenic, owing presumably to the great variety of aspects through which the aggregation passes is very much more prolonged. The synaptenic is succeeded by the pachytenic, where the filaments are coarser, then by the diplotenic, and this in its turn by the dictyate condition, which is the typical nuclear appearance of the young ovum. These changes pass gradually outwards, and by the tenth day even the cells quite at the periphery have passed through the earliest phases, whilst the central cells are reaching the final ones. By the fourteenth day there are a certain number of dictyate nuclei towards the centre. At this stage the central mesoblastic core is becoming obliterated, the egg-clusters of either side of the ovary very nearly meeting. The clusters are much smaller, having been split up by the ingrowing mesoblast. The number of dictyate nuclei now increases at a surprising rate, there being a great number by the fifteenth day, and a still greater number by the sixteenth day, by which time the clusters are almost indistinguishable, except round the periphery, their place having been taken by dictyate cells, some of which now show a surrounding follicular epithelium. There are also collections of small more or less rounded cells lying in between the young follicles, but not forming any part of the follicular epithelium. These are the interstitial cells of the ovary, and I propose now to trace their origin in detail. 1905.| the Interstitial Cells of the Ovary in the Rabbit. 39 Origin of Interstitial Cells—By about the tenth day the ovogenetic processes in the central egg-clusters are at their height, and continue in this condition, passing through the various phases, until about the fourteenth day, when, as already stated, some have reached their final stage. Through- out the whole period, however, there are in almost all the clusters some three or four cells, or perhaps more, which remain in the ordinary deutobroque state, and do not undergo any of the ovogenetic phases, the number of these being greater in the peripheral clusters than in the central ones; by the fifteenth and sixteenth days, when the number of dictyate nuclei is increasing, there are few, if any, of the deutobroque cells to be found in the central parts, but instead, there is an increasing number of the small round cells already referred to. As the days pass on the number of the former decreases, and the number of smaller cells increases. Thus, there is throughout the ovary, but in different parts at slightly different periods, a reduction in the number of deutobroque cells, which have remained unchanged, and an increase in the number of small round cells. Examination of sections of the fourteenth, fifteenth, and sixteenth days near the centre of the ovary, leaves no room for doubt that some of these cells form the follicular epithelium, gradually passing towards and arranging themselves around the young ova. At the fourteenth day the number of these cells to be found near the centre is not nearly sufficient to form the follicular epithelium for the large number of young ova, while near the periphery there are many more than would appear to be necessary for tne requirements of this part. By the fifteenth day the number of these cells near the centre has increased very largely, still more so by the sixteenth day, by which time many of the young ova are surrounded with follicle cells, and there are also the collections of these cells already referred to. Their number has meanwhile diminished somewhat at the periphery. The appearance in the intermediate parts gives the key to the whole question. Here are seen large numbers of these cells streaming inwards from the periphery and making their way between the ege-clusters of the periphery towards the centre, where the cluster formation can now be scarcely recognised. Here they arrange themselves around the young ova, or pass into little groups by themselves. These groups are the first beginnings of the real interstitial tissue of the ovary, and mark the commencement of the adult aspect of the organ. There are thus two main points to be emphasised at this period in the life of the ovary. First, the passage inwards of a large number of cells from the periphery, and secondly, the commencement of the adult formation by the formation of young follicles, and the appearance of interstitial cells. 40 Miss J. E. Lane-Claypon. On the Origin, etc., of [June i6, This passage inwards of cells from the periphery was noticed by Biihler ; he realised that the number of small cells near the centre during the height of ovogenesis was not enough to provide a follicular covering for all the young ova which were there, and he describes the streaming inwards of the cells from the periphery, and their passage to the young ova, around which they arranged themselves, and formed the follicular epithelium. Balfour noticed that in the later periods of ovogenesis there were present too many of the small cells, like the follicle cells, for them all to become arranged around the ova and give rise to the follicular epithelium. He was at a loss to account for the destiny of these supernumerary cells, and supposed that they must either eventually become ova or follicle cells, or be used up as food-stuffs for the other cells. It seems to me, however, that these cells, supernumerary as far as the follicular epithelium is concerned, are in reality very important. They form the groups which represent the interstitial tissue of the fully-formed ovary, and thus, far from being unimportant, are absolutely essential for the performance of the functions of the ovary. The question which now arises is, where do these cells come from, and what is their history of formation ? It has already been indicated that the number of unchanged deutobroque cells varies inversely with tle number of these cells, since these last are greater in number in the region where there are most deutobroque cells present, namely, at the periphery, especially in the region of the poles, and the high power of the microscope reveals the fact that these cells are indeed metamorphosed deutobroque cells. The ordinary deutobroque cell presents one or two irregular chromatin masses, from which pass out filaments of varying degrees of coarseness and fineness with nodules at their intersections. The centre of the nucleus is Clear, whilst around the periphery is the granular appearance already described. See fig. 2(/). In the ovary of about the eighteenth day the only regions where these cells are to be seen in any appreciable numbers are round the periphery and at the poles. They stand out even under the low power on account of their general transparent aspect as compared with the surrounding cells, and also in many cases on account of their rather larger size. There are also cells whose transparency is not so great, but which show up quite markedly in contrast to the rest and are rather smaller in size than the more transparent ones. These cells are transformation stages between the deutobroque and the ordinary interstitial type, and the process resembles very much in the inverse order that which has been already described for the deutobroque formation from the protobroque. The first stage is the gradual massing of the chromatin into irregular masses 1905.|] the Interstitial Cells of the Ovary in the Rablit. Al and the thinning of the chromatin strands, which become rather less in number, as do also the nodules (fig. 4(c)). At the same time the granular appearance extends gradually towards the centre, although it is not until quite a late stage that it reaches the centre itself (fig. 4 (f) and (g)). The size of the cell becomes gradually less, and the amount of protoplasm relatively greater. The retraction, as it were, of the filaments and strands towards the chromatin masses is very much more marked in the cells of smaller size, where there is a tendency for the masses to pass towards the periphery, leaving the centre clear (fig. 4 (d) and (e)). These changes continue until nearly all the chromatin is massed, the masses becoming rounder as the process goes on. There are always traces of strand formation left, in contrast to the protobroque nucleus, where it is markedly absent. Thus the small cell derived by differentiation from the deutobroque cell does not return to the characteristic protobroque type, but shows traces of its intervening deutobroque condition in the shape of strands of chromatin, and nodules on the strands. See fig. 4 (g). The cells, once reduced in size, become true ovarian cells, and may either function as follicle cells or as interstitial cells (fig. 4 (2) and (£)). Thus we find the following processes taking place in connection with the formation of the mature ovary. The cells of the germinal epithelium become embedded in the underlying mesoblast, and, once there, may either undergo differentiation, or apparently may remain in the protobroque condition. If the former be its fate, it must undergo nuclear transformation, together with growth in size, until it reaches the deutobroque stage. Arrived at this condition it probably divides, although possibly this is not an essential, and the two cells formed by this division are of the same type. There are now two courses open for the cells thus produced ; they may undergo the nuclear transformations of ovogenesis, and become primordial ova, or they may rest for a time, and finally undergo regressive transformations, becoming either follicle cells or interstitial cells. Every cell of the germinal epithelium is probably a potential ovum, relatively very few remaining in the protobroque state, although some may still be seen at the periphery in ovaries of the eighteenth day. Incomparably the greater number pass to the deutobroque state, preparatory doubtless to the formation of ova. All cannot become ova, for the other forms of cell are necessary for the maintenance of the ovarian functions ; possibly, therefore, only the most robust cells, and those which are most conveniently situated for obtaining nourishment undergo the ovogenetic changes. This would seem to be borne out by the fact that many more of the central cells, which are nearer their food supply, undergo ovogenesis, than 42 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, of the peripheral ones. The rest of the cells which are not able, for one cause or another, to undergo these changes, appear to remain quiescent for a while, until finally they regress and pass into a condition of subserviency to the needs of those which have become ova. Both follicle cells and inter- stitial cells are, however, still potential ova. They have passed through the initial stages, and only need enlargement and nuclear transformations in order to become ova, should the appropriate stimulus be given. This chance is not given to the follicle cells. As soon as the follicle begins to grow, they multiply rapidly, and probably provide, by their disintegration, the follicular secretion upon which the ovum feeds and grows. In the ripe follicle of the rabbit there is almost complete disintegration of the membrana granulosa, and the remains of the discus proligerus is presumably extruded with the ovum, perhaps serving it as food material prior to its fertilisation, and subsequent attachment to the uterine wall. The interstitial celis, however, have possi- bilities before them, being still capable of carrying out any function belonging to the true ovarian cell. All the true ovarian tissue is derived from the germinal epithelium, this tissue forming in the adult rabbit by far the greatest part of the whole ovary. There is relatively little mesoblast, which subserves solely the function of support and of nutriment-carrier to the rest of the organ. We may, therefore, look upon the whole ovary as consisting of two classes of cells and of two only, namely, (1) those derived from the germinal epithelium and performing all the ovarian functions, and (2) those derived from the original mesoblast, which are supporting and vascular. There remains only one feature to be dealt with in the immature ovary, one that has already been described by Balfour, namely, the protoplasmic masses formed by the aggregations of young ova. In the ovary of the sixteenth day the ova are all separate, but a day or two later this is not the case. There are now a large number of these masses of various sizes. They appear to consist of two, three, four, or even five young ova, to judge by the number of nuclei seen, but it is impossible to distinguish any trace of cell-boundary between them. Balfour suggests that these may either form as many ova as there are nuclei, or that one ovum may develop at the expense of the rest. This last point of view appears to be the more probable. It is evident that the massing takes place subsequent to the formation of the young ova, since it is not seen until after the appearance of the ova, and it would appear rather purposeless if they merely separated again a little later on. Moreover, in these masses one or two of the nuclei often look as if they were disappearing by gradual dissolution, and it is, therefore, probable that they will all ultimately serve as food-stuff for the one ovum 1905.| the Interstiival Cells of the Ovary in the Rabbit. 43 whose condition happens to have been best, and will, therefore, survive in the struggle for existence. This cannibalism on the part of the young ovum is not surprising, if the life of an ovum be considered. It is really but the normal condition of the cell at all its stages of development ; it grows and fattens at the expense of other cells. In the young ovary it is starting its first stage of growth and must devour other cells; later on, when it grows during the growth of the follicle, it lives upon the follicle cells, and later still, when, after fertilisation, the ovum in its extended sense refers to the young fcetus, it lives on the material provided by the cells of the maternal organism. This massing of cells and subsequent demolition of some of them for the benefit of one will be again dealt with in connection with the ovary of the pregnant rabbit. Changes in the Ovary during Pregnancy. The young ovary, after the period when it has reached a stage where the general aspect is that of an adult ovary, enters upon a period of slow growth, during which there is a continual formation of a considerable number of follicles, which having reached a state of partial maturity then begin to atrophy and finally disappear, leaving only a faint trace of their former existence in the shape of a scar. Having reached sexual maturity, the ovary becomes subject to periodic influences, of the nature of which little, if anything, is known. According to Fraenckel (9), they are intimately connected with the hypertrophy of the mucous membrane of the uterus. The sum total of the influences at work results in the production of “heat,” which occurs in the rabbit about once a fortnight, but the external changes in the vulva by which this is judged take place very gradually, so much so, that in the spring and summer time, when breeding is most prolific, the adult rabbit is scarcely ever out of one or other stage of “heat.” It is fairly certain, therefore, that whatever changes may take place in the ovary during “heat,” the condition recurs too frequently for these to be very marked. This does not refer in any way to the formation of the corpora lutea of “heat,” which are, of course, very definite. It has recently been stated by Heape (10) that unless impregna- tion occurs the ripe follicles of “heat” do not burst, in which case, presumably, there can be no formation of corpora lutea. If this is the case it would seem that there can be no such thing as the corpus luteum of “heat,” and the changes in the ovary during this period must be considered to consist merely of those taking place before sexual maturity, only rather more marked, namely, the formation of follicles, but after puberty these 44 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, reach the ripe state, since they will burst if impregnation occur, whereas this is probably not the case in the immature ovary. There would then, on this view of the case, be no ovulation except in the impregnated rabbit. It is quite possible that the additional stimulus of impregnation may hasten the bursting of the follicle, but it seems somewhat unlikely that without impregnation there should never be ovulation. The changes resulting in the production of “heat” are obviously those preparatory to a possible pregnancy. Fertilisation appears to be in itself a stimulus, and sets up general hypertrophy of the entire genital apparatus, producing likewise an improved condition of the animal; as to the mechanism of the production of this hypertrophy, however, our knowledge may be said to be nil, and we are reduced to classifying the whole as the changes brought ubout by the stimulus of pregnancy. Naked Eye Changes——F¥raenckel describes and figures very accurately the naked eye changes in the pregnant ovary of about the fifteenth day in his paper on the function of the corpus luteum. These changes are very striking, and indicate in themselves some very definite alteration or increase in the function of the gland; apart from the formation of the corpora lutea, there is an immense increase in absolute size, the gradual occurrence of which will now be described. The ovary of the non-pregnant rabbit is a small yellowish body, lying on either side against the posterior abdominal wall, a little below the kidney. It is usually about 4 inch in length and thin, being slightly wedge-shaped in transverse section and rather pointed longitudinally at either end; upon its surface may be seen clear round spots, showing the locality of the larger follicles, some of which, if they are nearly ripe, may even project slightly from the surface. The bursting of the follicles and fertilisation lead to the formation of the corpora lutea, the so-called “true” corpora lutea of pregnancy, and the growth of these bodies during the early period are undoubtedly the most characteristic feature in the naked eye appearance of the ovary. If the pregnancy be one with a large number of foetuses, the ovary often looks gnarled, so large and numerous are the excrescences produced upon its surface by these bodies. If these, however, be cut off, and if the organ be carefully examined at about the fourteenth day, when the corpora lutea are at their maximum state of development,* it will be readily seen that the ovary itself has increased in size, quite apart from the formation of the lutein tissue. The whole gland has a more swollen and rather less compact aspect; it is larger both in length and girth, and the wedge-shape of the * Of. Fraenckel, loc. cit. 1905.| the Interstitial Cells of the Ovary in the Rabbit. 45 transverse section is less marked; there are also in many cases fewer follicles in an advanced condition than in the non-pregnant state. Just at this period the energies of the gland have apparently been directed rather to the formation of the lutein tissue than new follicles. From the fourteenth to the eighteenth day the corpora lutea remain at their maximum, and then begin to diminish rapidly in size. Instead of being very vascular whitish bodies, projecting in many cases to the extent of three-quarters of their whole extent beyond the surface of the gland, they gradually diminish both in size and vascularity, until by about the twenty-second day of pregnancy they are merely elevations on the surface, showing the faintest possible trace of vascularity; this diminution continues steadily until, a little while before birth, the locality of these striking features of the fourteenth day of pregnancy is only seen by the presence of an opaque whitish circular area upon the surface of the ovary. The changes are so marked that it is possible after a little experience to diagnose very approximately the previous duration of the pregnancy from the appearance of the corpora lutea. Whilst these external changes are taking place in the lutein tissue, the rest of the ovarian tissue has been also undergoing changes, which, if not so striking in appearance, are none the less evident. It has already been stated that the organ at the fourteenth day shows marked increase in size apart from the corpora lutea; whereas, shortly after this period, these bodies begin to diminish in size, the reverse takes place in the rest of the ovarian tissue; and whereas growth of the ovary as a whole has been slow up to the present, it now becomes rapid and continues until close upon the time of parturition. By about the eighteenth or twentieth day all trace of wedge-shape in cross-section has completely gone, and the organ is nearly circular, the girth is much greater, and this increase extends right up to the poles. These changes become more and more marked, until at about the twenty-sixth day the organ is well over an inch in length, sometimes about 1} inches, showing a proportionate increase in its other measurements, and having a shape very much resembling a spindle with blunted ends. The number of clear round spots has meanwhile been increasing rapidly, so that in the majority of cases the greater part of the surface is taken up either by them or by the round whitish patches, which mark the spots where the corpora lutea have been projecting above the surface. The formation of follicles appears to be somewhat inhibited during the rapid growth of the corpora lutea, but to be resumed with greater energy when these have reached their maximum development. At the time of parturition there are a large number of 46 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, follicles which have almost reached full maturity, and it is a well-known fact that rabbits can be readily fertilised immediately after parturition. The gland, although soft, is not in any way brittle, and in spite of its great general enlargement retains on the whole the same shape, the most marked change being that from the wedge-shaped to the circular transverse section. Changes in Size of the Interstitial Cells——This great increase in size must be the result either of a large numerical increase, or of a very great increase in the size of the individual ovarian cells. The latter is at any rate the main, if not the only factor concerned, the change in size of the cell under the microscope being so marked as to attract attention even apart from any actual measurement. The measurements were made with a micrometer eye-piece, gauged against a micrometer slide, this method being found quite sufficiently accurate for the purpose. It was not intended to record exactly the size of each individual cell, but rather by taking the measurements as accurately as possible of a large number, to find the average increase in size at different stages of pregnancy. In taking the measurements considerable selection was exercised in the cells measured; only those whose area in section was approximately circular, and where the nucleus was centrally situated being used, as it was hoped by these precautions to obtain measurements passing as nearly as possible through the centre of the cell. The measurements are given below of 10 cells from each date of pregnancy, but this does not by any means represent the number actually measured, but the same figures recur again and again, and the average works out to almost precisely that given. Towards the end of pregnancy there is considerable difficulty in finding the right kind of cells to measure, nearly all of them being angular and irregular in outline, giving as a whole somewhat the appearance of a tesselated surface. The changes in general aspect of the sections, produced by the change in the size of the cells, will be returned to later on. 1905.| the Interstitaal Cells of the Ovary in the Rabbit. . 47 Interstitial Cells of a normal Rabbit. Diameter at 14th day of pregnancy. Diameter in mm. 00162 0:0180 00171 0:0189 00180 00153 00180 00198 00171 0:0189 01773 Average = 0:°0177 mm. Diameter at 18th day. mm. 0:0252 0:0279 0:0288 0:0279 0:0270 0:0288 0:0270 00252 0:0270 0:0279 0:2727 Diameter at 22nd day. mm. 0:0324 00315 00324 00315 00315 00306 00360 00351 00342 00315 0:3267 Average ="0°0272 mm. Average’ = 0:0326 mm. = 32°6'u mm. 0:0225 00216 00245 0°0207 00252 00216 0:0225 0:0270 0:0252 0:0243 0°2349 Average = 0:0234 mm. 23°4 w Diameter at 20th day. mm. 00542 0:0333 0:0306 0:0506 0:0297 00315 0:0324 0°0315 0°0324 0:0333 03195 mm. 0:0306 0:0288 0:0297 0:0342 0:0270 0:0333 0:0360 0:0351 0:0315 0:0324 03186 Average = 0:0319 mm. = 21D jn Diameter at about 26th day. Average = 00318 mm. = 318 p 48 Miss J. HE. Lane-Claypon. On the Origin, etc., of [June 16, Diameter just before birth, at 28th or Diameter a few hours 29th day. after parturition. mm. mm. 0:0288 00270 0:0315 0:0279 0:0270 0:0279 0:0297 00270 00288 Average = 0:0298 mm. 00270 Average = 0:027 mm. 0:0315 = 2058 0:0261 = 2 0:0288 0:0252 0:0306 00279 00315 0:0270 00306 0:0279 0:2988 i 0:2709 Diameter after Diameter after 3 days’ lactation. 6 weeks’ lactation. mm. mm. 0:0270 0:0180 00252 00171 0:0279 00162 00270 00180 00270 Average = 0:0268 mm. 0:0171 Average = 00171 mm. 0:0252 == ASM 00153 gg ILO 0, 0:0279 00180 00261 00171 0:0270 , 00162 0:0279 00180 0:2682 01710 Tabulating these results, one gets— Approximate age of pregnancy. Diameter of cells in p. OX(= normal) Weer eee e eee EO) TA tho daly. cceuAvanssespossene sasenneeee 23°4 Sth. se eaceeccee cere capensis: 27-2 QO th....,, wien eerdeaoneucemee seep ese Bley VAIS GRRE ESTO 3 Sexe, Seki oop OSB AGUS OP 32°6 Bit, 5. palevoge ace ee eee reer ae 31°8 ShortlivsbetonesoinhineeeeetPeereeees 29°8 ‘3 aTGCI A ye nian ween nae 27:0 3 Cans PDE SG aRARN ENE 26°8 6 weeks ,, s lc ill esl Rae al eal 1905.| the Interstitial Cells of the Ovary in the Rabbit. 49 Taking the radius of the cells it is seen that the increase in its length in the cell during pregnancy is from 8°5 to 16°3 or very nearly double. If the volume of the sphere be taken as 4 7 . r? and the cell be taken as a sphere, the ratio of the non-pregnant cell to the cell at the maximum size attained during pregnancy becomes almost exactly 1:7, which would allow sufficient enlargement of the ovary to account fully for the increase in size. It is not to be supposed that all the cells enlarge to the same extent, but it may reasonably be supposed that they enlarge to about five times their normal size. This will account for the enlargement of the whole ovary, and there would seem therefore to be no necessity to seek any further cause of the enlargement of the ovary during pregnancy. The only other possible cause which suggests itself at once is of course the division of cells, but although I have examined some hundreds of sections of pregnant ovaries, I have not found any trace of this happening. In giving the above figures I do not wish to suggest that the measurements are absolute. They are subject most probably to individual variations, depending possibly upon the number of fcetuses in each pregnancy, and on various other circumstances. The ovaries in question were, however, taken quite haphazard in regard to all external causes, which allows some scope for differences in the ovary, and the results are fairly definite. They show a great increase in the size of the ovarian interstitial cells during pregnancy, and that the main increase is reached by about the twenty-second day, and is sustained until just before birth, when there is a slight diminution in size. In this connection there is one feature to be dealt with, namely, the shape of the cell. Up to about the twenty-fourth or twenty-fifth day it is not difficult to find approximately spherical cells to measure. After this period, however, the difficulty of doing so becomes very great, if not impossible. The cells are angular and seem crushed together, and I would suggest that possibly the cells may be really still undergoing slight increase in size, but that the capsule having almost reached its maximum stretching capacity does not admit of the desired expansion, and the cells instead of being spherical become more closely packed in order to find room for the additional bulk, filling in as it were the interstices rather than causing au increase in size in the spherical direction. The rounded appearance is resumed very shortly after birth, and there is also a slight decrease in size. Why there should be a decrease before birth is a point upon which I feel it is impossible to offer any suggestion. The mechanism of the production of labour is a question upon which very little is definitely known; if, however, it be the function of the ovary to cause the adhesion of the foetus to the uterine wall (Fraenckel), a function carried out VOL. LXXVII.—B. iE 50 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, presumably by means of the interstitial cells, since these probably furnish the internal secretion of the organ, it seems not impossible that the diminution in size may be indirectly connected with the onset of labour. The Formation of “ Primordial” Ova from the Interstitial Cells. In addition to the increase in size there are other changes taking place in some of the interstitial cells near the peripheral parts of the ovary, during the later third (approx.) of the period of gestation. It is a matter of common histological knowledge that over the surface of the ovary there is a layer of epithelial cells, roughly about two cells deep, although varying slightly in thickness at different places. Immediately below this is a layer of tissue in which are embedded the primordial ova in their early stages, when they have not yet acquired a follicular epithelium or when that epithelium is not very highly developed. There are in addition groups of small ovarian cells which will eventually, as occasion arises, form the follicle cells for the primordial ova. This whole layer together with the germinal epithelium varies very considerably in thickness in different animals, the variation having possibly some relation to the age of the animal under investigation (cf. v. Beneden). In the non-pregnant animal and in the early periods of pregnancy, there is a fairly sharp boundary between these outer layers and the deeper lying interstitial cells. By about the twentieth day of pregnancy this state of affairs is seen to be gradually changing, and some of the interstitial cells are becoming surrounded by the connective tissue of the inner layer and thus getting cut off from their fellows below. Whether this is brought about by the passing outwards of the cells themselves or by the growth inwards of the connective tissue is very difficult to decide quite satisfactorily ; but I think it is reasonable to suppose that both processes are involved. It has already been shown that there are two means whereby the germinal cells of the embryo become embedded in the subjacent mesoblast, namely, by an ingrowth of the germinal cells and by a simultaneous upgrowth of the mesoblast lying below. Here we have an analogous condition, but the positions are reversed; the germinal cells are now inside and the mesoblast outside. This process, which is beginning to be evident about the twentieth day, continues throughout the rest of pregnancy, so that as the days go on more cells become cut off and press outwards, in many cases reaching almost to the periphery. The number of cells thus cut off varies appreciably in different animals, probably depending upon the age of the animal, but it is not excessive at any time; I have never found more than three or four rows 1905.] the Interstitial Cells of the Ovary in the Rabbit. 51 of cells cut off, and these rows do not form continuous layers round the ovary (vide fig. 5). About the twenty-third or twenty-fourth day, and in ovaries of later dates of pregnancy, a somewhat striking feature about some of these cells is that they are no longer mononucleated; two nuclei are frequent, three quite common, whilst in some cases there may be as many as six. These nuclei are not massed together as in a giant cell, but lie separate in the cell protoplasm. The latter is very much greater in amount than in an ordinary interstitial cell, and is irregular in outline. The appearance of these multinucleated cells suggests that they have been derived from the fusion of the same number of interstitial cells as there are nuclei in the cell. It will be remembered that van Beneden pointed out this appearance in the bat’s ovary, when he found in some cases as many as eleven nuclei, and he suggested that possibly one of them grew at the expense of the others, whom it used as food, or that one might become an ovum and the others the follicle cells. Examination of a large number of these cell masses shows that in many cases there is undoubted atrophy of one or more of the nuclei going on. In some there is a clear space where a nucleus might have been expected, in others the nucleus stains very faintly or only in parts, whilst there is usually one nucleus which stains intensely, especially in the iron hematoxylin specimens, and in which the staining, even after extreme differentiation, is still so dark as to remove all possibility of tracing any nuclear structure. This points to some difference of metabolic condition, and the conclusion seems obvious that this nucleus is growing strong at the expense of the others ; one is reminded of the protoplasmic masses described by Balfour in the young ovary and to which reference has already been made in this paper. Here we have a number of potential ova (for the fact has already been emphasized that all interstitial cells being derived from the germinal epithelium are potential ova) massed together, of which the nucleus of one of them grows at the expense of the others, which it uses as food material ; in the young ovary the end-product is a primordial ovum. In the pregnant ovary the end-product is likewise a “primordial” ovum. The cells of these aggregations are all quite clearly ordinary interstitial cells, and the surviving cell is also an interstitial cell differing only in the intensity of its staining reaction, It has I hope been conclusively shown, in the earlier part of this paper, that the interstitial cells have all been derived from the cells of the germinal epithelium, and have all passed through the deutobroque condition, and it has been pointed out by v. Winiwarter that if there is to be ovogenesis subsequent to the first great ovogenetic period, the cells which are to become E 2 52 Miss J. EH. Lane-Claypon. On the Origin, etc., of [June 16, ova must pass through the requisite nuclear changes. Also it is obvious, although this is not a point which he brings out, that there is a very great difference in size between the primordial ovum and the interstitial cell in a non-pregnant animal, and it is therefore necessary for the cell to enlarge at some period of the transformation. This requirement is fulfilled, as has already been shown, in the case of the pregnant ovary. The interstitial cell of the non-pregnant ovary has approximately a diameter of 17 y, but increases up to 29 w or even rather over 30 ~ in the pregnant animal. The size of a primordial ovum before it begins to grow, preparatory to becoming a Graafian follicle, is very constant; I have taken measurements of a large number of ova both in the young ovary and in the pregnant as well as the non-pregnant animal, and the average diameter is 27 yw, the diameter reached by the interstitial cells about the eighteenth day of pregnancy. It is about the twentieth day of pregnancy that the cutting off of the interstitial cells towards the periphery begins to be noticeable—that is to say, shortly after they have reached a diameter about equal to that of a primordial ovum. It is not, however, until a little later that the cells thus cut off begin to show any nuclear differentiation, in fact this is perhaps best seen in the ovary of a rabbit whose time of parturition has almost arrived. These changes are identical with those taking place in the deutobroque cells of the young ovary during the period of ovogenesis. The only difference lies in the fact that whereas in the pregnant ovary the process is taking place only at the periphery, and in relatively very small numbers, in the young ovary there may be 20 or 30 nuclei undergoing changes in the same field. The fact of their presence at all in the pregnant ovary is, however, all proof that is necessary for the formation of ova. It is not for a moment to be supposed that any formation of fresh primordial ova after the first great period should take place to anything like the same extent. Probably the actual changes only occur over a period of a few days, commencing about the twenty-fifth day of pregnancy, or rather earlier, and extending probably to a little after parturition. In the young ovary the changes do not commence until after birth, and some of the cells have completed their changes by about the tenth or eleventh day, the process being probably considerably less lengthy than this for the individual cells, and taking still less time, if any- thing, in the pregnant rabbit, where there is obviously a state of stimulation during the whole period of pregnancy. The first change passed through by the nucleus of an interstitial cell, which has passed to the periphery in order to become an ovum, is shown in fig. 6 (1). The nucleus shows chromatin filaments, in the middle of which are seen irregular lumps of chromatin. (In the diagrams the analogous stage of the 1905.| the Interstitial Cells of the Ovary in the Rabbit. 53 young ovary has been given side by side with that in the adult, and does not call for any special description.) Thisis a transition form from the interstitial nucleus to the leptotenic stage in the process of ovogenesis, and appears to be brought about by the breaking up of the nuclear chromatin into an immense number of filaments. The arrangement of the chromatin in the interstitial cells is, as a rule, discrete either in a rather loose reticulum or round the edges, usually the former. The first change is therefore the formation of fine filaments. The leptotenic stage of v. Winiwarter is brought about by the enlargement of the nuclear area and the spreading out of the filaments over this increased space, thus producing a looser arrangement which consists of fine filaments with a rather nodular appearance where they intersect (fig. 6(2)). This state would appear to be a very fugitive one (as observed likewise by v. Winiwarter), judging by the rarity of its occurrence. It is quickly passed through, and the nucleus enters upon the synaptenic condition (fig. 6 (3)). This stage occupies much longer than the last, and a relatively large number of nuclei are found in this condition, which has many modifications. The filaments at the leptotenic stage are spread out over the nuclear area, whilst at the final synaptenic the chromatin is massed into a lump at the side of the nucleus. All stages may be traced both in the adult pregnant ovary and the young ovary, but only the most characteristic phase is figured, namely, that where a very appre- ciable amount of massing has already proceeded, the mass being connected to the sides of the nucleus by a few very fine filaments. The massing completed, there seems to be a rearrangement of the chromatin, and it becomes spread out again, but this time the filaments are thicker. This is the pachytenic stage (fig.6(4)). The number of nuclei found in this stage is less than in the synaptenic, but still there are a fair number in various conditions. The filaments are so markedly thicker and more bulky generally that it is impossible to confuse it in any way with the leptotenic phase. The chromatin does not fill the nucleus quite so much as in the young ovary, but I have found sections where this was more the case than in the one figured; moreover in some the chromatin seems to have a more continuous disposition than is here represented. The transition stage between the pachytenic and dictyate or final stage is not, according to my observations, quite analogous to v. Winiwarter’s, and I rather hesitate to call it diplotenic, as the duality of the filaments is not well marked (sce fig. 6 (5)); the chromatin is still arranged in thick strands, and there is some trace of nucleoli, whilst at the same time there are a very few thinner nodulated strands, foreshadowing the condition called by v. Winiwarter dictyate, and which represents that of the young ovum. 54 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, The nucleolus in the dictyate condition (fig. 6 (6)) is very definite, and the chromatin is arranged more or less all over the nuclear area (which is now very large), and shows a number of small nodules both at what appear to be free ends and at the points of intersection. There can, in fact, be not much doubt that the changes taking place are identical with those seen in the young ovary, which lead to ovogenesis, and therefore it would appear that ovogenesis also takes place in the adult animal during pregnancy. Previous observers on this subject appear to have all considered that formation of ova must be accomplished by means of fresh invaginations of germinal epithelium, and those who thought they saw invaginations con- cluded at once that there was therefore a formation of ova in later life, whilst those who failed to find them denied the possibility on this account. My observations show that fresh invaginations of the germinal epithelium are not a necessity, but that the “invagination” has taken place already in the embryo. The invaginated cells of the germinal epithelium give rise to all the cells of the true ovarian tissue, which are all capable of functioning in any true ovarian capacity—that is, they may become ova or follicle cells, or interstitial cells, and most probably also lutein cells, their destiny appearing to be a matter of chance. The interstitial cells, however, are still capable of becoming ova, and of undergoing the changes requisite for ovogenesis should the appropriate stimulus be given. This stimulus is supplied when the animal becomes pregnant, and the ovarian cells enlarge in size. Towards the end of the time of pregnancy some of them press towards the periphery and undergo the necessary changes, becoming true ova. Thus every pregnancy would seem to be a stimulus for the next, in the way of providing new ova, although even of the relatively small number found probably very few ever reach maturity. Conclusions—Summing up the conclusions reached in this paper we find— 1. That a large number of germinal cells become embedded in the subjacent mesoblast. Of these the great majority undergo transformations up to a certain stage. This stage having been reached, they may pass through the necessary processes of ovogenesis, or they may become modified to form either follicle cells or interstitial cells, this last process being the chief fate of the cells near the periphery, whilst ovogenesis is that of the more centrally situated ones. 2. The interstitial cells are thus potential ova, capable of becoming ova should the appropriate stimulus be given. 3. This stimulus is provided by pregnancy, during which period the interstitial cells undergo enlargement in size, exceeding that of a primordial ovum. 1905.] the Interstitial Cells of the Ovary in the Rabbit. 55 4, About the twenty-third day some of the interstitial cells become cut off near the periphery and pass through the nuclear transformations of ovogenesis, becoming true ova. I wish to express my deep obligation to Professor Starling, under whose supervision this research has been carried out, and without whose never- failing assistance, interest, and sympathy at each step of the work it would have been impossible to carry out the investigations described above. Also I desire to thank Mr. H. G. Plimmer for his kindness in giving me much valuable information in regard to the carrying out of the histological details. BIBLIOGRAPHY. Amann, ‘ Verhandlung der Ges. Deutsch, Naturf. und Arzte in Miinchen,’ 71 Vers., vol. 2, 1899. 2. Ancel and Bouin, ‘Comptes Rendus Soe. de Biol.,’ 1903 and 1904. 3. Andrews, ‘ Journ. of Obst. and Gyn. of British Empire,’ May, 1904. 4. Balfour, ‘Quart. Journ. Micros. Science,’ 1878. 5 6 = Van Beneden, ‘ Arch. cle Biol.,’ vol. 1, 1880. . Born, ‘ Arch. f. Anat.,’ 1874. 7. Biihler, ‘Zeit. f. Wissen. Zool.,’ 1894. 8. De Foulis, ‘Quart. Journ. Micros. Science,’ 16, ser. 3, 1876. 9. P. Fraenckel, ‘ Arch. f. Gynik.,’ vol. 68, part 2, 1903. 10. Heape, ‘ Roy. Soc. Proc.,’ April, 1905. 11. His, ‘ Arch. f. Mikros. Anat.,’ vol. 1, 1865. 12. Koster, ‘ Meded. der Kon. Akad. der Wetenschappen,’ 2 reek, teel 3, 1868 13. Macleod, ‘ Arch. de Biol.,’ vol. 1, 1880. 14. Nagel, ‘ Arch. f. Mikros. Anat.,’ vol. 31, 1888. 15. Pfliiger, (1) ‘ Ailgem. Medizin. Centralzeitung,’ 1861 and 1862 ; (2) Uber die Eierstécke der Saugethiere und des Menschen,’ 1867. 16. Schrén, ‘Zeit. f. Wissen. Zool.,’ vol. 12, 1863; ‘Inaug. Dissert. der Med. Facultat zu Erlangen,’ 1862. 17. Wagener, ‘Arch. f. Anat. u. Phys. Anat.,’ Abthg., 1879. 18. Waldeyer, ‘ Hierstock und Ei,’ Leipzig, 1870. 19. v. Winiwarter, ‘ Arch. de Biol.,’ Liége, 1900. 20. Wendeler, ‘ Martin’s ‘Die Krankheiten der Eierstécke und Hier,’ Leipzig, 1899. For further literature on the early history of the ova see :— Von Baer, ‘De Ovis Animalium et Hominis Generi,’ 1827. Barry, ‘ Phil. Trans.,’ London, 1838. Billroth, ‘ Miiller’s Arch.,’ 1856. Bischoff, ‘ Bau des Menschlichen K6rpers,’ Sémmering, 1842. Borsenkow, ‘ Wiirz, Naturwissen. Zeitschrift,’ vol. 4, 1863. Coste, ‘ Embryogénie Comparée,’ p. 80. De Graaf, ‘De Mulierum Organis Generationi Inservientibus,’ 1668. Grohe, ‘ Virchow’s Arch.,’ vol. 26. 56 Miss J. E. Lane-Claypon. On the Origin, etc., of [June 16, Henle, ‘Handbuch der System. Anat.,’ vol. 2, 1866. Kolliker, ‘ Entwickelungs-geschichte des Menschen und der Hoheren Thiere.’ Langhans, ‘ Virchow’s Arch.,’ vol. 38. Letzereich, Pfliiger’s ‘Untersuchungen aus dem Physiol. Lab., Bonn,’ 1865. Purkinje, ‘ Encyclopidisches Wérterbuch,’ vol. 10, Artikel “ Hi,” Berlin, 1834. Romiti, ‘Arch. f. Mikros. Anat.,’ vol. 10, 1874. Rouget, ‘Comptes Rendus de I’Acad. des Sciences,’ 1879. Spiegelberg, (1) ‘ Nachricht d. k. Ges. der Wissenschaft,’ Géttingen, 1860 ; (2) ‘ Virchow’s Arch., vol. 30, 1864. Stricker, ‘Wiener Akad. Sitzungsbericht. Math. Naturw. Klasse,’ 2 Abthg., vol. 54, 1866. Valentin, ‘ Miiller’s Arch.,’ 1838. R. Wagner, ‘ Beitrige zur Geschichte der Zeugung und Entwickelung, p. 42. Wharton Jones, ‘London and Edinburgh Phil. Mag.,’ No. 39, Sept. 1835, p. 209. DESCRIPTION OF PLATE 1. Fie. 1.—Ovary of 20th day embryo (rabbit). Fixed in Gilson’s fluid :— @ = protobroque cells. b= ; (in mitosis). ce = deutobroque cell. d = connective tissue cell. e = position of mesoblastic core. Fie. 2.—Cells in ovaries of rabbits just before and after birth. Fixed in Gilson’s fluid :— a = protobroque cell. b | | ‘ > = transition forms between (a) and (/). e J = deutobroque cell. Fic. 3.—Ovary of three days old rabbit showing formation of egg-clusters. sublimate solution :— a = protobroque cell. 6 = deutobroque cell: C= - in mitosis. d = leptotenic stage in ovogenesis. Fie, 4.—Cells in ovaries of young rabbits. Fixed in Gilson’s fluid :— @ = deutobroque cell in ovary of three days old rabbit. b= % 3 about 18 days old rabbit. ¢ | , = transition stages from deutobroque to ovarian cell in ovary t | 18 days old rabbit. 9g 3 A = interstitial cell = Policia cell } trom ovary of young rabbit (18 days). Fixed in of about Lane-Claypon. Roy. Soc. Proc., B. vol. 77, Plate 1. 1905.| the Interstitial Cells of the Ovary in the Rabbit. 57 Fie. 5.—Ovary of rabbit about 22nd day of pregnancy. Taken from the cortical region. Fixed in sublimate solution :— a = germinal epithelium. 6 = primordial ovum. e = multi-nucleated interstitial cell. d = interstitial cell, becoming isolated. f = connective tissue. g = modified germinal cells. Fie. 6.—1—6 are taken from the ovaries of rabbits in the later stages of pregnancy. la—6a from ovaries of young rabbits : showing ovogenetic changes for comparison with 1—6. Fixed in Gilson’s fluid :— 1 = transition from interstitial to leptotenic phase in pregnant ovary. la = ” 9 young 9 2 = Leptotenic phase in pregnant ovary. 2a = ” yoans ” 3 = Synaptenic phase in pregnant ovary. 3a = i young % 4 = Pachytenic phase in pregnant ovary. 4a = ” young " £h) 5 = Diplotenic phase in pregnant ovary. 5a = 9 young os 6 = Dictyate phase in pregnant ovary. 6a = ” young ” 58 Ferivlity in Scottish Sheep. By Francis H. A. MarsnHatt, M.A. (Cantab.), D.Sc. (Edin.), Carnegie Fellow, University of Edinburgh. (Communicated by Professor EH. A. Schafer, F.R.S. Received August 10, 1905.) My attention was first directed to the subject of fertility by Mr. Walter Heape, to whom I am much indebted. Experiments have been described in agricultural publications on the effects of different methods of feeding and general treatment upon wool or meat production; but excepting, so far as I am aware, for Mr. Heape’s report on “ Abortion, Barrenness, and Fertility” in sheep in the South of England for the year 1896 to 1897,* no systematic attempt has been made to deal with the factors which influence fertility either in the sheep or in other animals. Numerous experiments, however, are annually conducted by flock-masters for a practical object, and it has been thought desirable to put the results of some of these on record with a view to making comparisons, and in the hope eventually of reaching definite conclusions upon this subject. That differences in food and environment exercise an influence over fertility in the sheep as in other animals has long ago been recognised,+ and recently attention has been called to the wide range of variability in the sheep’s sexual capacity, this animal showing a complete gradation between the moncestrous condition and the most extreme degree of polycestrum.} Asa preliminary step in an investigation on fertility in sheep, it was decided to issue a schedule of queries addressed to various flock-masters chiefly in the East of Scotland. The present communication consists of a condensed account of some of the information contained in their replies.§ The preparation and issue of the schedule was undertaken by the Highland and Agricultural Society of Scotland, under whose auspices the work is being carried on. I am under no light obligation to the members of this society for their co-operation, as well as to all those gentlemen who have supplied * “ Abortion, Barrenness, and Fertility in Sheep,” ‘Journ. Roy. Agric. Soc.,’ vol. 10, 1899. ‘Note on the Fertility of Different Breeds of Sheep,” ‘ Roy. Soc. Proc.,’ vol. 64, 1899. + Darwin, “ Animals and Plants,” Popular Edition, London, 1905. Marshall, ‘The Cistrous Cycle and the Formation of the Corpus Luteum in the Sheep,” ‘Phil. Trans.,’ B, vol. 196, 1903. § It is hoped that a full report may be issued next year in the ‘Transactions’ of the Highland Society. Fertility in Scottish Sheep. 59 me with information regarding their personal experiences. The com- paratively small number of schedules issued and returned (the latter being about 50), rendered it possible to obtain fuller information than would other- wise have been the case, while the information obtained in this way was in some cases supplemented by personal conversation or further correspondence. For the purpose of showing the percentages of lambs per ewes, of barrenness, and of abortion (Tables I, II, and IIL) among flocks treated in different ways, these are divided into six groups as follows :—* Division A.—This includes hill sheep (Scotch Black-faced and Cheviots), which were kept all the year round on the sides of hills, and received no sort of special treatment. Dwision B.—This includes hill sheep (Scotch Black-faced and Cheviots), which were placed upon better grass at tupping time (7.¢., during the sexual season) or shortly before. Division C._—In this division are included half-bred (Border Leicester x Cheviot) ewes which underwent a process of flushing by being fed on turnips, cabbages, oats, dried grains, maize, or other artificial food during tupping and for about three weeks before. The ewes were in most cases merely fed on grass during the greater part of the year, but received a certain amount of extra food (turnips, ete.) during the latter part of pregnancy (7, usually from about the beginning of the year). Division D.—This includes two flocks of Cheviot ewes which were flushed at tupping time but were fed on grass during the rest of the year. Table I—Number of Lambs per 100 Ewes. locks: Under | 90 | 100 | 110 | 120 | 180 | 140 | 150 | 160 | 170 | 180 | 190 Total. 90 p.c.| p. c. | p.c. | p. ¢. | p. ¢.| p. c.| p. c. | p. c.| p. c. | p.c.| p.c.| p. c Division A 2 10 2 1 — | — | — | — | — | — | — |] KH 15 noe BL = Li hop Sf} dp ee tS ee pee Sf 4 5 GC] = —{;—|;—j]1 i | = | 2 3 i | =} 2 10 ye. oD: — — | — | — 1 1 — | — | — | — | | 2 » = — | — | — | = | =| 2 1 Bi} al 1 7 »p of a ee SS 1 2 2)/—|]— 6 Total...... 2 10 3 1 3 2 3 4 8 3 1 3 44, iT { The numbers represent the numbers of flocks, the total being 44. The percentages are the percentages of lambs per ewes in the different flocks. The flocks are arranged in six divisions, according to the methods of feeding, as explained in the text. * The variation in the number of flocks in the three tables is due to the flock-masters not having supplied complete information im all cases. Some flocks, therefore, are included in one table but not in another. 60 Dr. F. H. A. Marshall. [ Aug. 10, Table I1.—Percentage of Ewes that Aborted. Flocks. | None. ae lp.c. | 2p.c | 3p.c. | 4p.c. | Sp.c. | Total. Division A ......... — — 6 3 2 2 iL 14 Bias ESS ed vane a 3 == = = TL a 4 is nal OP Aas 4 3 2 1 = ce 10 Bee aD ecitccne — | 1 = _— — — 2 Sy iets! Ohnescmeneee 2 3 1 1 = = 7 op We yrecsiatasene 2 1 1 = — 1 _— 5 Total .-....- 8 11 11 5 | 2 4 1 42° | The numbers represent the numbers of flocks, the total being 42. The percentages are the percentages of ewes which aborted in the different flocks. The flocks are arranged in six divisions, according to the methods of feeding, as explained in the text. Table IIJ.—Percentage of Barren Ewes. Flocks. None. oa lp.c.| 2p.c.|3p.c | 4p.c | 5p.c | 6p.c. | 7p.c. | Total. Division A| — ay —_ 2 3 — 5 2 iL 14 eBid = us 1 1 1 we il ae 4 we 3 — — 3 1 1 — — — 8 Sad ae 1 1 _ a0 = es as 2 eh ge 2 4 — a —-};-}] =— 7 eS es = 2 1 ee 1 1 = = 5 Total...... 3 1 5 12 5 4 6 3 1 40 The numbers represent the numbers of flocks, the total being 40. The percentages are the percentages of barren ewes in the different flocks. The flocks are arranged in six divisions, according to the methods of feeding, as explained in the text. Division H.—This includes flocks of Border Leicester and half-bred (Border Leicester x Cheviot) ewes which were placed on better pasture durig tupping and for some time (usually about three weeks) before, but which otherwise received no sort of special treatment; in some instances, however, the ewes received a limited number of turnips during pregnancy. Diwision #.—TVhis division includes Border Leicester, and half-bred (and a few Cheviot) ewes which were fed all the year round on grass, receiving no special treatment of any kind. Table I shows very clearly that the percentage of lambs was, as a rule, larger among flocks which underwent a process of artificial stimulation during the sexual season, while Table III shows that the percentage of barren ewes was generally relatively less in such flocks. The Cheviot and Black-faced sheep in Division B which produced less than 100 lambs per 100 ewes (Table I) are stated to have been unusually unprolific owing to their never 1905. ] Fertility in Scottish Sheep. 61 having properly recovered from the extreme cold in March and April, 1904. This case, therefore, may be regarded as exceptional. The percentage of barren ewes in this flock was six (Table ITI). In the three cases in which the percentage of lambs was over 190 the exact numbers were 191-5 per cent., 193°75 per cent., and 196 per cent. In the first of these the ewes (which were half-bred Cheviot x Border Leicester) were fed on grass only, during the previous summer. For three weeks (during tupping) they were given a full supply of turnips on grass, and between tupping and lambing (five months) they were given a mixture of dried grains and turnips, and “lamb food” for three weeks before lambing. The rams (which were pure Border Leicesters) were given bruised oats during tupping. No record was kept of the ages of the ewes. One ewe had four lambs and 12-5 per cent. had triplets. In the second case the ewes (half-bred) were fed upon Bombay cake, bruised barley and a little linseed as well as turnips and cabbages during tupping (after grass), and some turnips were given during pregnancy. The rams (Border Leicester and Oxford Down) were similarly treated. The ewes were all three-shear. Flushing with turnips was found to bring the ewes in season very rapidly. Triplets were produced by 13:5 per cent. of the ewes. The third case is recorded under Division E, but ought possibly to have been included under Division C. At tupping time the ewes (which were half-bred) were put upon better pasture, and between tupping and lambing they were given some turnips and as much cut hay as they would eat. Previously to tupping they were fed on grass alone.* The ewes were all ages up to four-shear. The rams (which belonged to the Border Leicester, Oxford Down and Cheviot breeds) were supplied with no artificial food at tupping. The twins appear almost invariably to have been born early during lambing time, thus showing that the reproductive activity of the ewes is generally greatest early in the tupping season. Only two returns record that twins were mostly born late, while 28 state that early twins were the rule, both among the artificially fed flocks and those which received no special treatment. There is abundant evidence also that flushing hastens forward the tupping time. It has recently been shown that “heat” in animals is almost certainly brought about by an internal secretion elaborated in the ovaries. It would appear, therefore, that the artificial feeding exercises a stimulating influence * Cheviot ewes, kept on the same farm, and treated similarly, produced only 10:0 per cent. lambs. + Marshall and Jolly, “ Contributions to the Physiology of Mammalian Reproduction. Part Il.—The Ovary as an Organ of Internal Secretion,” ‘ Phil. Trans.,’ B, vol. 198, 1905. 62 Dr. F. H. A. Marshall. Fertelity in Scottish Sheep. over the secretory activity of the ovaries, while at the same time causing the Graafian follicles to mature more rapidly and a larger number to discharge during the earlier cestrous periods in the sexual season. Regarding the effects of artificial feeding during one tupping season upon the fertility of the sheep in after years, it has so far been difficult to obtain precise information. The opinion usually expressed is that flushing is not detrimental to subsequent fertility unless it is overdone; but in a very few of the returns the view is stated that the after-effect is adverse. It is also said that if ewes are flushed one year the process must be repeated the next: otherwise the ewes tend tv be less fertile than if they had never been flushed at all. On the other hand, several of the returns show that sheep which produce twins one year very frequently bear twins also in the year following. This seems to occur irrespectively of whether it was the practice to flush the ewes. It would appear, therefore, that an increased degree of fertility is characteristic of certain particular ewes. That fertility is a character which can be inherited admits of no doubt. It is to be noted, however, that with the breeds considered in this paper, twius are seldom if ever selected for purposes of tupping, since they generally are not so well developed, owing to their having had less nourishment when they were young lambs. It would seem, therefore, that the fertility of these breeds is diminished owing to the fact that the rams which are probably naturally the most fertile are the ones which are the least frequently employed for breeding 63 On the Nature of the Galvanotropic Irritability of Roots. By AurreD J. Ewart, D.Sc., Ph.D., F.L.S., and Jessie §. Bayuiss, B.Sc. (Communicated by Francis Darwin, For. Sec. R.S. Received September 7,— Read November 23, 1905.) After the contradictory statements of Elfving* that roots curve towards the positive electrode (anodotropic), and of Muller-Hettlingen,t that they were kathodotropic, Brunchhorst{ apparently reconciled these contradictory observa- tions by finding that strong currents, like those used by Elfving, produced a curvature to the positive electrode, weak ones a curvature to the negative electrode. The former curvature Brunchhorst considered to be traumatropic in character, on the ground that it was shown by decapitated roots, whereas the negative curvature was not. The proof that the galvanotropic irritability resides solely in the root tip, is, however, quite insufficient, and hence Brunchhorst’s conclusion does not appear to be justified by the facts. The methods of the first two investigators leave much to be desired, and although Brunchhorst’s experiments were, in part, carried out on a klinostat, they are by no-means perfect. Thus the roots were immersed in water in a closed vessel, through which the current was passed by means of carbon electrodes. Apart from the effects due to the gases occluded by the electrodes, and to the deficiency of oxygen in the water, there would always be a tendency for the current to run obliquely or longitudinally through the roots, whose tissues form better conducting media than the surrounding water. This tendency will be especially pronounced when the roots are not exactly at right angles to the current, as is practically always the case, and when, as in Brunchhorst’s experiments, numerous roots are examined at the same time. Finally, although Brunchhorst gives some data as to the total amount of current flowing in the circuit, these data afford no evidence as to the actual amount of current passing through the individual roots. Evidently, therefore, the supposed positive and negative parallelo-galvanotropism of roots is by no means satisfactorily established, and accordingly Miss Bayliss undertook to reinvestigate this subject, under more well-defined and controllable conditions, and with the results given in brief below.§ The strength of constant current required to produce a curvature is incredibly small, for using a voltage of approximately 1:3 volts, a resistance * ‘Bot. Zeit.,’ 1882, p. 257. + Miiller-Hettlingen, ‘ Pfliiger’s Archiv,’ vol. 31, 1883, p. 193. { Brunchhorst, ‘Ber. d. D. Bot. Ges.,’ 1884, vol. 2, p. 204. § Full details will be given by Miss Bayliss in a later paper. 64 Dr. A. J. Ewart and Miss J. S. Bayliss. On the [Sept. 7, of 100,000 to 150,000 ohms was required in the circuit, so that the current passing through the 1 to 3 sq. mm. of cross-section lay between 0°0000135 and 0:000009 of an ampere. Even then it was difficult to produce a eurva- ture without serious injury, or even fatal effects in the case of sensitive roots. When the platinum electrodes were on opposite sides of the apex, the curva- ture was always towards the positive electrode. If, however, one electrode was placed on the non-irritable base of the root and the other to one side of the apex, the curvature always took place towards the current side, inde- pendently of which electrode was on the apex. These results were obtained upon a klinostat into which the current was led by mercury contacts, and transmitted by platinum electrodes to the stimulated region of the root. The seedling and wires within the rotating glass cylinder were insulated on a slab of paraffin wax. The facts observed suggested that the curvatures were not the result of any parallelo-galvanotropic irritability, but were due to the accumulation of the products of electrolysis at the points of application of the electrodes. Confirmation was obtained by exposing the roots to strong currents (voltage 1 to 4) for short periods (five to eight minutes), and then rotating on a klinostat, when exactly similar results to the above were given. Furthermore, if the anodal region was cut out of an electrolysed root and applied to one side of the apex of another, a curvature was shown to this side. In addition, the application of minute squares of absorbent paper, moistened with decinormal acid or alkali, caused curvatures towards the stimulated side, whereas ordinary neutral paper produced no effect in air saturated with moisture. When the acid and alkali were applied simultaneously on opposite sides, the curvature always took place towards the acid side. This corresponds to the curvature towards the positive (acid) electrode produced by moderately strong currents. The weakest currents used produced similar positive curva- tures, and hence Brunchhorst’s negative curvatures cannot be explained by Weber’s law, as being due to the normal acidity of the root tissues preventing the stronger stimulating action of the acid coming fully into play until it accumulates beyond a certain limit. The curvatures are usually completed in from 6 to 24 hours after exposure to .the current, but they may be distinctly perceptible within four to six hours, and may begin in one to two hours, under optimal conditions. Hence it is not surprising that if the roots are fixed in a plaster cast after stimula- tion, and rotated on a klinostat for one or two days, a rapid sharp curvature is produced on freeing the root from the cast, whereas after two to four days the effect of the stimulation has passed away. All of these curvatures can be produced without any of the cells of the root being killed, and even when an 1905.|] Nature of the Galvanotropic Irritability of Roots. 65 injury is produced, the curvature is usually towards the injured side, instead of away from it, as in a true traumatropic curvature. The curvatures produced by continuous currents appear usually to be accompanied or preceded by a temporary more or less pronounced retardation of the average rate of growth inlength. Indeed the latter may be temporarily arrested for some time after strong stimulation, even when the electric current produces little or no injury. In such cases negative results may be obtained as regards curvature. Finally, using non-polarizable electrodes moistened with cell-sap diluted with distilled water, no curvatures were produced, whereas similar stimulation, using platinum electrodes applied to the surface of the root, and with the non-polarizable electrodes still in the circuit so that the resistance was the same, gave the usual curvatures according to how and where the electrodes were applied. With stronger currents and more prolonged exposure, curva- tures are induced, even when “ non-polarizable” electrodes are used, since the products of electrolysis may diffuse to the surface of the root, and it is impossible to prevent the internal polarization which takes place wherever the current traverses dissimilar saline solutions separated by semi-impermeable membranes. There is, however, less tendency to injury than with platinum electrodes. The irritable and responsive zone extends 4 to 5 mm. behind the apex of the rootof Vicia Faba and Phaseolus vulgaris. When one platinum electrode was applied to the non-irritable base of a root, and the other laid flat on the extreme tip, no curvature was produced in whichever direction the current was passed. This is presumably due to the products of electrolysis diffusing evenly and stimulating the irritable regions and cells equally on all sides, for when the same current was applied transversely behind the apex, a positive curvature was shown. If the roots were either truly positively or truly negatively parallelogalvanotropic, they should curve in the above experiment so as to place the tip parallel to the current, and either against or with its direction, whenever this does not at first coincide with their tropic irritability. The “ galvanotropism ” of roots is therefore due to chemotropic stimulation by the products of electrolysis, of which the acid is more effective than the alkali, the latter also being neutralised more or less by the respiratory carbon dioxide. It is indeed possible that the curvature of the roots of Lupinus albus in gelatine towards phosphates and carbonates observed by Lilienfeldt* may be of similar origin, since acid phosphate and alkaline carbonates were used. That the “galvanotropic” or galvanogenic curvatures are not trauma- * Lilienfeldt, ‘Ber. d. D. Bot. Ges.,’ 1905, vol. 23, p. 91. VOL, LXXVII.—B. F 66 Mr. F. Keeble and Dr. F. W. Gamble. Jsolation of [Oct. 6, tropic in origin is shown by the fact that they may be produced without any cells being killed. In Brunchhorst’s experiments the electrolysis presumably occurred in the superficial cells of the roots submerged in water, the tissues being sufficiently impermeable superficially to the liberated acid and alkaline ions to allow them to accumulate beyond the minimum for stimulation. Although the curvature is usually sharp and strongly localised to the point of application of the electrode, the discriminatory power of the root, as well as the relation of the rates of growth on concave and convex sides to the normal rate of growth, suffice to show that the response is a stimulatory one, and is not due to the direct action of the products of electrolysis, retarding growth on one side or accelerating it on the other. On the Isolation of the Infecting Organism (“ Zoochlorella”) of Convoluta roscoffensis. By FREDERICK KEEBLE, M.A., University College, Reading, and F. W. GAMBLE, D.Se., University of Manchester. (Communicated by Sydney J. Hickson, F.R.S. Received October 6, 1905.) The present paper gives a preliminary account, (1) of experiments proving that the green cells (“‘zoochlorelle”) of Convoluta roscoffensis result from infection from without: (2) of the means whereby the infecting organism may be cultivated outside the body of the animal: and (3) of the nature of the infecting organism. 1. Evidence for Infection—In our former papers* we reached the con- clusion that though direct proof of infection was lacking, the evidence pointed most strongly to infection as the source of the green cells of Convoluta. We showed, moreover, that the difficulty in the way of obtaining direct proof of the origin of these green cells is due to the fact that the mucil- aginous capsules that invest the clutches of eggs laid by Convoiuta are rarely, if ever, sterile. Even when adults are washed repeatedly in sterilised sea- water and caused to lay in sterilised surroundings, their egg-capsules become covered in time with a varied flora of colourless and of green organisms. It is therefore necessary to isolate the young at the moment of hatching. During the present summer we have done this in larger numbers than before and maintained them in carefully filtered sea-water. Such young Convoluta * “The Bionomics of Convoluta roscoffensis,’” ‘Roy. Soc. Proc.,’ vol. 72, p. 93, and ‘Quart. Journ. Micro. Sci.,’ vol. 47, p. 363, 1903. 1905.| the Infecting Organism of Convoluta roscoffensis. 67 remain colourless and may be kept in this condition for at least a month without showing any sign of infection, whilst at any time batches of them may be caused to become green in one to three days by the addition of sea- water or of cultures of the infecting organism. 2. The Cultivation of the Infecting Organism.—All attempts to cultivate green cells taken from the body of Convoluta have failed. Haberlandt made an unsuccessful attempt, we ourselves were equally unsuccessful, and so also was Miss Harriette Chick, who brought to the task great experience of such researches and the most recent methods. The problem had therefore to be attacked from the other end. If the green prisoners of Convoluta never escape alive, the only chance of obtaining the infecting organism lies in catching it before its entrance into the animal. A scrutiny of many attempts to obtain colourless Convoluta in large quantities revealed the fact that generally egg-capsules, isolated in sterile water, give rise to Convoluta which remain colourless for a fairly well-marked period of two or three weeks. After this time, however, green specimens make their appearance. Sometimes the number of green animals thus appearing is few; more often it increases with great rapidity. Such results suggested that the infecting organism occurs sporadically on or in the capsules; that it divides freely in this situation; and that after a period of vegetative division it is liberated in sufficiently large numbers to infect the hundreds of Convoluta experimented upon. During the past summer this hypothesis has been put to the test and found to be correct. Large numbers of egg-capsules were kept in filtered water and the young Convoluta upon hatching were removed, so that the vessels contained only empty capsules or capsules the eggs of which had failed to hatch. These vessels were kept under observation. At the end of three weeks several minute spherical bodies of a spinach-green colour were detected. Upon microscopical examination these proved to be colonies of green cells enclosed by and filling an egg-capsule. During examination the membrane around such a colony bursts, and the contents, previously quiescent, swarm out of the capsule, revealing themselves as so many unicellular flagellated green organisms. It remained to apply the infection test. Samples of colourless Convoluta reared in sterilised surroundings were put into the vessel containing these flagellated cells. They became infected, and in the course of two or three days exhibited in their tissues green cells identical in character with those of normal Convoluta roscoffensis. Similarly, sterilised sea-water containing cultures of these green organisms is as potent as ordinary unsterilised sea-water in producing infection. F 2 68 On the Infecting Organism of Convoluta roscoftensis. Ordinary sea-water or such cultures alike induce infection within a few days, when added to just hatched Convolutas reared previously under sterile conditions; whereas the addition of sterile sea-water to samples from the same stock of Convoluta produces no infection. 3. Nature of the Infecting Organism—Whilst reserving for a detailed and illustrated account the full description of the organism infecting Convoluta roscofjensis, we may here briefly refer to its more salient characters and systematic position. In its adult and holophytic stage, the motile green organism is ovoid and flattened in front. At the anterior end it possesses two pairs of similar flagella. A basin-shaped chloroplast envelops the greater part of the body, and is turned in anteriorly, forming a clear border to a colourless funnel-shaped area which runs axially for a third of the cell’s length. A plate-like red “eye-spot” or stigma, with a darker rim, lies somewhat excentrically a little in front of the middle of the cell ana not in relation to the flagella. The resting nucleus is spherical. Near the posterior end of the chloroplast is a large octagonal pyrenoid, provided with a starch sheath, giving it a somewhat irregular outline. True starch, giving a marked blue colour with iodine, is present in quantity. A cell-wall is either absent or of extreme tenuity in the organism when first liberated, but later on a marked wall of mucilaginous character is demonstrable and may reach a considerable thickness. These characters indicate that the green cells of Convoluta roscoffensis are true alge, belonging to the Chlorophycee and allied to Chlamydomonas. The presence of four equal flagella suggests that they belong to the genus Carteria.* * Blackman and Tansley, ‘New Phytologist,’ vol. 1, p. 23, 1902. 69 Further Observations on the Germination of the Seeds of the Castor Oil Plant (Ricinus communis). By J. REYNOLDS GREEN, Sc.D., F.R.S., Professor of Botany to the Pharma- ceutical Society of Great Britain, and Henry Jackson, M.A., Fellow and Tutor of Downing College, Cambridge.* (Received March 22,—Read May 18, 1905.) About 15 years ago one of the authors carried out a series of researches on the germination of the seeds of the castor oil plant (Ricinus communis),t and endeavoured to ascertain the course of the decomposition and utilisation of the reserve materials which are present in the seed. As the results of this investigation formed the starting point of the present series of researches, it will be well at the outset to restate the conclusions which were then arrived at. The larger part of the reserve materials of the seeds of Ricinus, which are laid up in the cells of the endosperm, consists of the well-known castor oil. The amount varies in different seeds, but it ranges from as little as 50 per cent. to upwards of 80 per cent. There is a considerable amount of proteid matter in the cells, most of which is found in the so-called aleurone grains. These have a somewhat intricate structure; an ovoid mass of phytoglobulin, soluble in 10 per cent. solution of common salt, surrounds a proteid crystal, soluble in saturated solution of the same salt. In the grain by the side of the crystal there is a rounded aggregation of mineral matter, the so-called globoid, long considered to be a double phosphate of calcium and magnesium, but probably a more complex body containing its phosphorus in some form of organic combination. According to Vinest the proteids of the grain are an albumose and a globulin; in the opinion of Osborne and Harris§ this is not the case, only a globulin being present, probably identical with the edestin of the hemp seed. There are other substances present in small amount, but in very trifling pro- portions when compared with the oil and the proteids. * The present series of experiments was commenced by me in collaboration with Mr. W. T. N. Spivey, of Trinity College, Cambridge. After his lamented death in 1901, Mr. Jackson took his place.—J. R. G. + Green, ‘Roy. Soc. Proc.,’ vol. 48 (1890),. p. 370. t Vines, “ Proteid Substances in Seeds,” ‘Journ. of Physiol., vol. 3 (1880), p. 91. § Osborne and Harris, “ Nitrogen in Protein Bodies,” ‘ Amer. Chem. Journ., vol. 25 (1903), p. 335. 70 Prof. J. Reynolds Green and Mr. H. Jackson. [Mar. 22, The conclusions arrived at in 1888 with reference to the changes set up during germination were the following :— “1. The reserve materials in the endosperm of Fucinus communis consist chiefly of oil and proteid matters, the latter being a mixture of globulin and albumose. “2. The changes during germination are partly due to enzyme action, there being three enzymes present in the germinating seed: one is a protease resembling trypsin, the second splits the oil into fatty acid and glycerine, the third is a rennet enzyme. “3. At least two of these, and therefore presumably all of them, are in a zymogen condition in the resting seed, and become active in consequence of the metabolic activity set up in the cells by the conditions leading to germination, especially moisture and warmth. “4, The changes caused by the enzymes are followed by others, due to the metabolism of the cells, these being processes of oxidation. “5, The embryo exercises some influence on the latter, setting up as it develops a stimulus probably of a physiological description. “6. The result of these various processes is to bring about the following decompositions :— “The proteids are by the enzyme converted into peptone, and later into asparagin. “The oil is split by the glyceride enzyme into fatty acid and glycerine ; the latter gives rise to sugar, and the former to a vegetable acid which is soluble in water and in ether, is crystalline, and has the power of dialysis. “7, Absorption in all cases takes place by dialysis. “8. The appearance of starch and of oil in the embryo or the young plant is due to a secondary formation, and not to a translocation of either.” FORMATION OF LECITHIN. The advances in our knowledge of the metabolic processes of plants that have been made during the interval that has elapsed since the publication of this paper, and the new methods of experiment that have been introduced, suggested that the work which was admittedly incomplete and tentative should be taken up again. There remained especially the question of the meaning of the reserve supplies of phosphorus and the part which they take in the general metabolism accompanying germination. The aggregates of phosphates referred to as the globoids of the aleurone grain undergo a change during the process, by virtue of which they slowly pass into solution. 1905.| On the Germination of Seeds of the Castor Oil Plant. 71 As this change supervenes upon the development of an acid reaction in the seeds, it seems not unlikely that it may be caused by the action of the organic acid which is formed in the cells of the endosperm almost as soon as germination begins. On resuming the work a more careful examination of the oily contents of the endosperm cells led to the discovery that they contained, mixed with the oil, a certain quantity of a substance, the decomposition products of which pointed to its being a lecithin (a peculiar fatty body containing phosphorus). The resting seeds were pounded in a mortar till they formed a homogeneous paste. This was extracted for some hours with ether in a Soxhlet’s apparatus, and was afterwards twice extracted further on a water-bath with absolute alcohol, the flask being fitted with a reflux con- denser. The alcoholic and ethereal extracts were mixed and evaporated to dryness at a gentle heat on a water-bath, and the fatty residue fused with dry carbonate and nitrate of potassium until all trace of free carbon had disappeared. After cooling, the fused residue was dissolved in water, and the addition of ammonium molybdate and nitric acid produced a yellow precipitate, indicating the presence of phosphorus. The quantity of the latter was ascertained by converting it into magnesium pyrophosphate and weighing. A little of the oil was then hydrolysed by boiling with baryta, when there separated out a flocculent precipitate of a barium salt, which, after washing and drying, was found to contain no phosphorus. This barium salt had the characteristic soapy appearance of the stearates. It was found possible to identify cholin in the endosperm of the germinating seeds, as will be more fully shown a little later (p. 74). The only way of estimating the amount of the lecithin present in the alcohol-ether extracts of the endosperm was to determine accurately the phosphorus as magnesium pyrophosphate. Great precautions were taken to ensure the repeated use of very dry ether so as to exclude the possibility of extracting any inorganic phosphates. A little of the oily residue from the extraction was incinerated in a platinum dish, and it was found to leave no inorganic ash. We therefore assumed that all the phosphorus extracted as deseribed was originally present in the complex lecithin form. Taking the formula usually given for lecithin (CssHaNPO,), we calculated the amount of the latter that would be present. Its average amount was equal to 0°236 per cent. of the dry weight of the seeds. Several series of experiments were made to investigate the changes in the fatty constituents of the endosperm during germination. The action of the fat-splitting enzyme known now as lipase was confirmed, and the early 72 Prof. J. Reynolds Green and Mr. H. Jackson. [Mar. 22, stages of germination were found to be as set out in the former paper. Consequently, examinations of the contents of the seeds were made at certain stages of the germination, and before the process had begun. The stages were the following :—(1) The seed at the time of the cracking of the testa, usually after 24 to 48 hours in the soil; (2) the seed with the radicle protruding for a length of 1 to 2 cm., usually about three days after sowing; and finally, (3) seeds whose lateral root system had become fairly well developed. The times at which these stages were reached varied with the samples of seeds used, and the temperatures at which the germination took place. The results of a typical experiment are stated in the subjoined table :-— Table A. Oil in seeds. Fatty acid in seeds. Dry Lecithin Degree of weight | per cent. development. of seeds Percentage | Percentage} of weight used. aa ral of weight Gi ee of weight | of seeds. WEE of seeds. | “eSB | of seeds. grammes. | grammes. grammes. Resting seeds ............ 4°48 3°7115 828 O1 2-2 0°236 | Seeds just cracking} 4°47 3 016 67-5 0-204 4°6 at | testa . Radicle protruding 1—2| 4:17 2-19 52°53 05 ibt) 0-478 cm. Lateral roots spreading. 3°34 0-789 23 ‘6 0-565 16 ‘89 0 873 Root system estab- lished It will be seen that the amount of lecithin diminished during the early stages of germination, the reserve supply becoming almost exhausted. After the young seedling had begun to develop, however, there was a gradual increase in the amount. This increase was maintained during the later stages and was fairly constant till the endosperm was used up. There was clearly a consumption of the oil throughout. The amount of lecithin, though small, varied somewhat in different experiments. In one series it was in much larger proportion than in that quoted. The residue soluble in alcohol and ether amounted to 0-9 per cent. of the weight of the, resting seeds, and in the later stages of germination the amount present rose to approximately 2 per cent. This quantity, however, in our experiments was exceptional. These experiments suggest that in the utilisation of the fatty reserves lecithin certainly plays a part and, possibly, a predominant part. 1905.| On the Germination of Seeds of the Castor Oil Plant. 73 Lecithin has been shown by Overton * to be a normal constituent of living cells, and to exercise considerable influence on the transport of various materials across the limiting layers of the protoplasm. It has no doubt also a certain, though at present undetermined, nutritive value. The composition of lecithin is indicated by the change which it undergoes on hydrolysis, when it is decomposed into stearic (or palmitic or oleic) acid, glycero-phosphoric acid and cholin. CuHoNPO, + 3H20 = 2CisH3602 + CsHoPOg + CsHisNO2. Lecithin. Stearic acid. Glycero- Cholin. phosphoric acid. From this, its constitution has been represented as J oC 0GiHss C3H;—-OCOC7Hs; : \O — PO(HO)0(CH2)2N(CH;),0H Only a trace of it exists in the resting seed; as it increases during germination and the quantity remains fairly constant during the whole period of absorption of the fatty reserves by the seedling, we have evidence of a formation of it during the germinative processes. The endosperm contains such substances as may yield the several groups necessary for its formation. The decomposition of the oil by the enzyme lipase can furnish the fatty component, belonging to the oleic group, and at the same time the glycerine of the glycero-phosphoric acid. The phosphorus of the latter is at hand in the shape of the phosphatic globoids whose solution has already been alluded to. The ‘nitrogenous body cholin may be looked for among the products of the decomposition of the proteids of the aleurone grains. Examination of the contents of the endosperms during germination ultimately established the presence of all these constituents. The fatty acids and the glycerine were identified in 1888, and the methods of detection and estimation were quoted in the former paper. A careful examination of the phosphates of the globoids, taken for purposes of comparison from seeds at the respective stages of germination quoted in Table A (p. 72) showed that their solution proceeded side by side with that of the oil. No change in them could be observed under the microscope till the testa was cracking, and the time of its inception varied a good deal. In the early stages, prior to such cracking, no reaction for phosphorus could be obtained from a watery extract. The quantity of phosphorus present in the resting seed was 0:205 per cent. of the dry weight; this diminished in Stages 2, 3, and 4 of Table A to 0:16, 0:14, and 0-11 per cent. The globoids are decom- * Overton, ‘ Pringsheim’s Jahrb.,’ vol. 39 (1900). 74 Prof. J. Reynolds Green and Mr. H. Jackson. [Mar. 22, posed gradually but fairly rapidly during the germination, and in the later stages contribute to the acidity of the cell-sap, which contains phosphoric acid. Search was made in a mass of endosperms for cholin. The germinating seeds were ground up in a mortar and allowed to stand for some days under alcohol which was nearly absolute. This was decanted and evaporated to dryness, the residue being again extracted with absolute alcohol and subsequently by a mixture of alcohol and ether. These extracts were mixed and evaporated to dryness, leaving a final residue, soluble in water. When a strong aqueous solution of this was boiled, decomposition took place, and a gas was evolved which possessed the well-known ammoniacal and fishy odour characteristic of tri-methylamine. The decomposition can be represented by the following equation :— CH.(OH)CH.N(CH;);0H = CH.(OH)CH,OH + N(CH;). When to some of an aqueous solution of the residue from the alcoholic and ethereal extracts a little platinic chloride was added, after standing for some time the characteristic yellow octahedral crystals of the compound which cholin forms with platinic chloride separated out. These were soluble in 15 per cent. alcohol, and on combustion yielded a residue of metallic platinum. We have thus all the constituents of lecithin present in the germinating seeds. It was difficult to apply the ordinary tests for lecithin when so large a quantity of oil was present. Towards the close of the germination, however, conditions were more favourable, the lecithin being present in relatively large proportion. The existence of a proteolytic enzyme of a tryptic nature in the germinating seeds was shown in the former paper. Among the products of its action a considerable quantity of crystalline amino-bodies were detected, though not sufficient for a complete analysis. They separated out from the concentrated alcoholic extracts, after removal of the sugars, in quantities that enabled their amino-nature to be proved. The power of the enzyme to produce these in vitro has already been noted.* We have found the cholin also to be due to the action of this enzyme. 150 cc. of an extract of the endosperms of a quantity of germinating seeds was prepared by steeping them for several hours in water containing 0°2 per cent. of formaldehyde as an antiseptic. It was then strained through muslin and filtered till it appeared as a clear straw-coloured liquid. This was divided into two, and half of it boiled ;to destroy the protease. A quantity of globulin was * Green, loc. cit., p. 377. 1905.] On the Gernunation of Seeds of the Castor Oil Plant. 75 prepared from a further quantity of the same germinating seeds by extracting them with a 10 per cent. solution of common salt, and precipi- tating the proteid by addition of alcohol. The precipitate was rapidly collected on a filter, washed and suspended in a little water. The 75 cc. of the extract that had not been boiled was put into a beaker and 5 ¢.c. of the suspended globulin added; a similar preparation was made of the 75 c.c. that had been boiled. Both were kept in an incubator at 40° C. for a week. At the end of that time digestion was complete in the unboiled preparation, the globulin having disappeared, leaving a morbid solution. Both were perfectly free from bacteria, the formaldehyde acting extremely efficiently as an antiseptic. The two digestions were then filtered and the filtrates evaporated to dryness. The residues were extracted successively with absolute alcohol and with a mixture of absolute alcohol and ether, each extraction .being continued for two days. The first alcoholic extract was evaporated to dryness and the residue again extracted with ether. The two ethereal extracts were subsequently mixed and evaporated to dryness and the residue taken up with a little water. There was considerably more of this residue in the digestion carried out by the unboiled extract of the seeds than in that associated with the other. To each a little platinic chloride was added in watch-glasses, and they were set aside. After 24 hours, in both cases minute crystals had settled to the bottom of the liquid, which were soluble in alcohol of 15 per cent. concentration. From this solution the characteristic yellow octahedra slowly settled out, and these gave the same reactions as those prepared from the extracts of the endosperms. The amount obtained from the digestion by the unboiled extract was much greater than that from the boiled one, though the latter yielded some, attributable no doubt to a certain quantity present in the 75 c.c. of the original extract of the seeds employed. The experiment shows, therefore, that the cholin of the lecithin can be prepared from the proteids of the seeds by an enzyme which is developed during germination, and is presumably the enzyme already described as a trypsin. The similarity of this enzyme to the trypsin of the pancreas is borne out by the occurrence of tryptophane among the products of its activity both in the plant and in vitro in the laboratory. The contribution of material for the synthesis of lecithin does not seem, however, to be the only result of the decomposition of the fat. There is not sufficient phosphorus in the resting seeds to enter into the composition of as much lecithin as the fat would produce. It is, of course, possible that the lecithin may be decomposed during consumption and part of its phosphorus 76 Prof. J. Reynolds Green and Mr. H. Jackson. [Mar. 22, set free to combine again, but even then the quantities do not seem to be proportional. Another fate must attend a considerable quantity of the fat. To this point we shall return later. These results suggest that the utilisation of the oily reserves is a much more complicated process than was supposed. The enquiry took from this point a wider range, and soon involved the abandonment of the idea that the separate reserves undergo independent changes during germination. THE SUGARS OF RICINUS. A more complete study of the sugar was next undertaken. Du Sablon showed, in 1895* that it is a mixture of at least two sugars, one of which has not the power of reducing Fehling’s solution. In our experiments, a large number of seeds having been germinated, the endosperms were separated from the embryos and ground to a paste in a mortar. The mass was then extracted with large quantities of water, by keeping it for some hours in a steriliser at 100° C., removing the water at intervals till the extract showed that all the sugar had been dissolved. The extracts were mixed and concentrated to about one-tenth of their volume. Addition of normal acetate of lead separated from this extract the acids present, together with the bulk of the proteids and certain other constituents. These were filtered off, and the sugars were precipitated from the filtrate by adding basic lead acetate and ammonia. The precipitate was separated by filtration and suspended in water, and the lead removed by a stream of sulphuretted hydrogen. The solution so obtained was concentrated, and the process repeated, the final solution being then concentrated to a thick syrup, which showed the presence of two constituents possessing different solubilities in alcohol. By a repetition of concentration and extraction, the syrup was ultimately separated into two parts, one of which reduced Fehling’s solution, while the other did not. Unfortunately the separation did not involve the complete isolation of the two sugars, as the reducing power of the first fraction was always increased after boiling with dilute mineral acid. The increase was not constant in different preparations, a fact which pointed to incomplete separation rather than to the reducing sugar being of the maltose type. The second fraction of the syrup was, however, free from the reducing sugar. Treated with invertase or with a dilute mineral acid it speedily reduced Fehling’s fluid. A quantity of it was concentrated nearly to dryness and with some difficulty dissolved in alcohol. Addition of ether to a little of the solution caused precipitation of the sugar. To the great bulk of the * Du Sablon, “ Sur la Germination des Graines Oléagineuses,” ‘Rev. Gén. de Bot.,’ 1895, p. 145. 1905.] On the Germination of Seeds of the Castor Owl Plant. 77 solution, therefore, a little ether was added, drop by drop, till a faint turbidity was apparent. After standing in this condition for some days, a crop of aggregates of erystals separated out. When dissolved in water they were found to have a specific rotatory power of about ap = +66. After inversion with a dilute mineral acid the specific rotatory power became about ap = —18. The solution of the crystals gave no crystalline osazone on warming with phenylhydrazine acetate. These reactions are fairly conclusive that the non-reducing sugar is cane- sugar. The reducing sugar was refractory and no method succeeded in rendering it crystalline. It was also found impossible to separate it completely from the cane-sugar, so that its specific reducing power could not be obtained. Readings with the polarimeter were unsatisfactory on account of its proving impossible to free its solutions from a yellow colouration. When the latter were warmed with phenylhydrazine acetate they yielded a quantity of a pale yellow osazone which analysis proved to be the osazone of a hexose. After several recrystallisations from alcohol and from ethyl acetate the crystals were found to have a constant melting point at 204°C. This is consistent with the view that it is invert sugar produced from the cane-sugar with which it is associated. It negatives the hypothesis pnt forward in the former paper that it is derived from the glycerine of the fat, for this sugar (glycerose), now much more completely investigated, is known to yield an osazone melting at 130° C. to 131° C.* The occurrence of two sugars exhibiting the characters just described suggested a search for invertase among the constituents of the endosperm. A good number of well germinated seeds were selected, having most of the endosperm absorbed; the embryos were well developed, their root system considerably branched. The endosperms were removed and ground up into a paste, which when strained through muslin yielded 95 c.c. of an acid sap. This was carefully neutralised and a little antiseptic added. It contained a quantity of both reducing and non-reducing sugar, 10 c.c. of the sap reducing 0-2 gramme of cupric oxide. Tubes were prepared containing respectively 10 c.c. of the neutralised juice with 10 c.c. of a solution of the non-reducing sugar from the seeds, and 10 cc. boiled juice with the same quantity of the sugar solution, and they were digested in a water-bath at 40° C. for several hours. On titration the weight of cupric oxide reduced by the digestion containing unboiled juice was 0°31 gramme while the other * Fischer and Tafel, ‘ Ber. d. deut. Chem. Ges.,’ vol. 20, p. 1088; Fenton and Jackson, ‘Trans. Chem. Soc.,’ 1899. 78 Prof. J. Reynolds Green and Mr. H. Jackson. [Mar. 22, gave the same weight as the original juice, 0°2 gramme. The treatment with the juice had increased the original reducing power 50 per cent., showing the presence of invertase. Further experiments upon the same point showed that invertase appears in the endosperms at a very early period of germination, usually after a few hours ; it is well established in 48 hours, and increases in amount up to the stage at which a good root system has been established. In a series of experiments upon its development during the germination three stages were compared : (1) The seeds had the radicle protruding about 0:3 inch; (2) The roots were 1 inch long and the secondary rootlets were just cracking the primary root; (3) There was a good root system and the endosperms were about half consumed. Extracts were made of all these and 2 c.c. of each allowed to act on 20 c.c. of a 1 per cent. solution of cane-sugar, at 40° C. for 24 hours. They were then titrated with Fehling’s fluid, when the weights of cupric oxide obtained were :— (1) 0003 gramme; (2) 0:006 gramme; (3) 0-007 gramme. These experiments lead us to the conclusion that the sugars of the endosperms may be put down as cane-sugar and invert sugar. The relative quantities of these two sugars during the progress of germina- tion have been ascertained and are given in Table B. Experiments on this point have been published by Du Sablon in the paper already referred to. He states that he found non-reducing sugar to be slightly in excess of reducing sugar in the resting seed and to increase more rapidly than the latter till the radicle is about 1-5 to 2 inches long, when the reducing sugar becomes equal in amount and, later on, preponderates considerably. Our experiments were carried out in the following manner :—A number of seeds were germinated in sawdust in an incubator kept at a temperature of 22°C. In each experiment three were taken, peeled, and ground up to a smooth paste in an agate mortar. The paste was then boiled with a sufficient quantity of water for an hour, the extract strained off, filtered, and divided into two. Half was warmed to 40° C. with 1 cc. of a solution of invertase prepared from yeast, and kept at that temperature for 24 hours. The invertase solution was ascertained to be free from sugar or other substance capable of reducing Fehling’s fluid. The two halves of the extract were then titrated side by side, and the weight of the cupric oxide taken in each case. From these weights the quantities of the two sugars were computed in the usual way. . 1905.] On the Germination of Seeds of the Castor Oil Plant. 79 Table B. Time of orb Invert Cane- germination Come thion Of aq eods wine sugar in sugar in in hours. ae milligrammes. | milligrammes. @) Resting seeds ............s00seseeeee | 11 10-7 45 Caruncle swollen ................- | 2-7 5°17 69 Little further external change...) 2°3 0) 117 Root about 0°75 inch long ...... | 6-7 19 -4 168 Root 1°35 inch long ............... 5-2 10°5 216 Roots branching .................5 | 19°65 35 °7 240 Endosperms cracking ............ | 29 -O1 35 °8 312 Good root system .............00665 | 40 °8 52 °6 | A comparison of this Table with Table A suggests that the course of events in which the sugars are involved proceeds upon much the same lines as that connected with the lecithin. The cane-sugar is present in greater quantity in the resting seeds, it gives place to invert sugar under the influence of the invertase during the early period of germination, and subsequently increases in amount and remains slightly in excess of the invert sugar during the later stages when absorption is more active. This suggests that cane-sugar is the actual reserve, and that the invert sugar represents the form which has the greater nutritive value. In accounting for the increase in the quantity of cane-sugar which marks the progress of germination, it is necessary to call attention to a fact noticed for the first time a few years ago by Mr. Biffen in the Cambridge Botanical Laboratory. Emphasis has already been laid upon the fact that a very vigorous metabolism in the endosperm cells is an accompaniment of germina- tion. This was commented on by Van Tieghem* in 1877, when he found that endosperms deprived of their embryos were capable of swelling and. apparently starting a kind of development. In the former paper on this subject one of us described experiments} confirmatory of Van Tieghem’s views. Biffen has found that a considerable increase of the protoplasm of these endosperm cells is a marked feature of the early stages of germina- tion. The exact time at which it occurs varies somewhat, but it corresponds fairly closely with the recommencing formation of cane-sugar. The coincident occurrence of these two events points to a vrowth of the protoplasm of the endosperm cells at the expense of the initial reserves, which we have seen are undergoing conversion changes at and before this time, and a subsequent construction of further carbohydrate reserves by * Van Tieghem, “Sur la Digestion d’Albumen,” ‘ Comptes Rendus,’ vol. 84, p. 578. + Green, loc. cit., p. 389. 80 Prof. J. Reynolds Green and Mr. H. Jackson. [Mar. 22, the protoplasm in the endosperm for the nutrition of the outlymg embryo as its growth continues. Apart from such secretion the endosperm contains no carbohydrate material, while the latter seems to be essential for the maintenance of merismatic tissue. The fact that this carbohydrate substance is cane-sugar coincides with the observation of Brown and Morris* that cane-sugar is always present in the growing embryo of the barley-grain. It appears to be a form of carbohydrate very suitable for serving as a temporary reserve material, more easily utilisable than starch, and therefore formed where the deposit of the reserve will be of very short duration, as in the case of the embryo, and in that of the foliage-leaf, where Brown and Morris found it at a very early period of the photosynthetic construction. Indeed, from the results of analyses of the mixed sugars then present they suggested that it might even be the first sugar formed.t It may again be noted that in the case of Ricinus its formation is accompanied or speedily followed by the secretion of invertase. The enzyme is not present in the resting seeds, but develops in the endosperms after exposure to a temperature of 25° C. in moist earth or sawdust for 48 hours or less, though germinative changes are not visible so soon in the external appearance of the seeds. The amount of the enzyme increases continuously all the time of germination, and the invert sugar increases coincidently. The protoplasm appears to keep up a secretion of cane-sugar and the invertase seems to keep working on the latter, so as to supply invert sugar at once to the protoplasm of the cells and to the young absorbing embryo. It will be seen from what has been said that we do not associate the formation of this carbohydrate material during the germination directly with the diminution in quantity of the oil which is taking place at the same time. Our experiments lend no support to the views of Sachs that the oil was directly transformed with either sugar or starch. The two processes are features of a new metabolism set up in the cells as germination becomes established. To this point we shall return later. THE ACIDS OF THE GERMINATING SEEDS. The question of the nature of the acid to which the reaction of the germinating seed is due remains to be dealt with. Evidence of acidity can * Brown and Morris, “Researches on the Germination of some of the Graminez,’ ¢ Journ. Chem. Soc.,’ vol. 57 (1890), p. 518. + Brown and Morris, “A Contribution to the Chemistry and Physiology of Foliage Leaves,” ‘Journ. Chem. Soc.,’ May, 1893, p. 673. 1905.| On the Germination of Seeds of the Castor Ou Plant. 81 be obtained after a seed has been exposed to warmth and moisture for 24 hours, and it becomes more and more intense for six or seven days. While the reaction to litmus paper becomes very prominent, only very small quantities of acid can be obtained from the seed. The expressed juice of a parcel of germinating seeds was titrated with decinormal potash solution, and 10 c.c. of it neutralised only 4 cc. of the alkaline solution. We made several attempts to prepare it in quantity by experimenting upon about a thousand seeds at once. They were germinated for a week, and the endosperms separated from the embryos, ground and boiled in water in a steriliser for several hours. After straining and filtering part of the extract was distilled by the aid of steam. The distillate was practically neutral in reaction, the merest trace of acidity coming over. The acid in the remainder, after removal of uncoagulable proteid, was precipitated by normal lead acetate, and the lead salt filtered off, suspended in water and treated with a stream of sulphuretted hydrogen till the lead was all con- verted into sulphide. The filtrate from the latter was concentrated to a small bulk, and the precipitation and subsequent treatment repeated. The final filtrate was concentrated to a small bulk in vacuo over sulphuric acid. The acid residue, somewhat syrupy in consistence, was then washed repeatedly with dry ether, which dissolved a certain quantity, leaving behind, however, a good deal of acid which was soluble in water only. The bulk of the latter was ascertained to be phosphoric acid. The solution in ether was concentrated i vacuo and formed a syrupy residue. We found it impossible to crystallise this acid or to obtain a crystallisable salt. Many attempts were made to effect crystallisation, but in only one case was any success obtained, and then only a few crystals on the surface of the syrup were formed. Unfortunately, therefore, the nature of the acid has not been ascertained. After looking for the source of this organic acid we again find reason to attribute it to the oil. We have already pointed out (p. 75) that the amount of lecithin formed is not sufficient to account for the disappearance of the whole of the oil of the seed, but that another fate awaits a considerable quantity. It was suggested in the former paper* that the acid of the germinating seed was derived from the oil by certain processes of oxidation, and served as the means of its utilisation. It is extremely unlikely that this acid is directly or indirectly connected with the sugars. We think we have here the explanation of the gradual diminution of the oil in the early stages | of germination, and of the development of the coincident acidity. The acid reaction of the endosperm sets in before any change can be detected in the * Green, loc. cit., p. 385. VOL, LXXVII.—B, G 82 Prof. J. Reynolds Green and Mr. H. Jackson. [Mar. 22, globoids of the aleurone grains and before any reaction for phosphoric acid is obtainable. The probability of an oxidation of the oil taking place in the early stages of germination has already been pointed out. This is now ‘rendered still more probable by the discovery of an oxidase in the germinating seeds. On mixing a strained and filtered extract of the endosperms with a solution of hydroquinone, the colour of the latter speedily becomes pink and, later, red. The extract gives instantaneously a blue colour with an emulsion of guaiacum, and slowly turns a solution of pyrogallol purple. Boiling the extract destroys the power of setting up these changes. The oxidase adheres very tenaciously to the tissue of the endosperm, and it is very difficult to extract it completely. Though the oxidase can be extracted and the extract found to act on such easily oxidisable bodies as those mentioned, no attempt has succeeded in making it oxidise ricinoleic acid outside the plant. This may, however, be due to non-attainment of the conditions which exist in the cells of the endosperms. ‘Though its appearance is suggestive, it has not been proved that it plays a part in the oxidative processes of the fats, if the latter take place. The probability of such oxidative processes is considerable, for, in addition to the considerations just put forward, it should be remembered that one of us has shown that the formation of the acid is dependent upon the access of oxygen. In seeds germinated in its absence, though part of the oil was transformed, no acid soluble in water was formed.* The problem is complicated by the fact that the distribution of the lipase, invertase, and oxidase of the germinating seed is practically the same. NUTRITION OF THE EMBRYO. The sequence of changes which has, so far, been described, suggests a modification of the views now current as to the mode of utilisation of reserve materials in albuminous seeds. It has been commonly held that the efforts of the parent plant ceases with the deposition of reserve food in or near the embryo, in such a condition as to be easily used. Possibly, also, certainly in some cases, the parent is responsible for the provision of an enzyme to effect the change of the reserve food into a suitable condition for absorption. The utilisation is, however, attributed more or less fully to the embryo. In many cases the latter secretes the enzymes itself, and in others it is the active agent in absorption. The metabolic changes in the endosperm attributable to the parent are held to be more or less independent of each other, and to consist of the enzyme actions only, each enzyme fitting its appropriate food for absorption. * Green, loc. cit., p. 389. 1905.| On the Germination of Seeds of the Castor Oil Plant. 83 This, as we have shown, is far from being the case with Aucinus. Here we have a series of most complex changes set up by the parent in the endosperm, accompanied by a renewed growth and revived secretory activity of the parent itself. The various constituents are made to act upon each other under the influence of the protoplasm of the endosperm cells, the latter showing a great increase in the amount of their protoplasm, while the protoplasm initiates a complex metabolism comparable in intensity with any which can be marked in the adult plant. It feeds itself, having prepared the food from the reserves ; it secretes new products, which were represented but sparingly in the original cell-contents, thus preparing a new and completely representative food supply which it places at the disposal of the embryo. At the same time, however, the latter plays a considerable part in the scheme of nutrition, besides carrying out the processes of absorption. A study of the distribution of the enzymes of the seed shows us that the preparation of food is not all carried out by the parent. The lipase was stated in the earlier paper* to originate in the endosperm cells and to continue to be developed there during the whole course of the germination. The invertase and the oxidase appear to have a distribution similar to that of the pase. The trypsin, however, originates in the embryo. In the course of the researches made by Mr. Biffen, which have already been referred to, he found that the epidermis of the young cotyledons contained cells, occurring at short intervals, which stained quite differently from the rest, and were full of granular contents. We prepared a large number of cotyledons from seeds in course of germination, taking them at an early stage when it was just possible to separate them cleanly from the endosperm. They were then washed carefully in warm distilled water till all organic matter was removed from their surfaces. Each cotyledon was then cut in half along the mid-rib. One set of halves was dipped for a moment in boiling water. The two sets were put into a solution of the globulin of the seeds prepared by dissolving it from the seed in 10-per-cent. solution of common salt and precipitating it by strong alcohol. The tubes containing them were put for a few hours into an incubator at 30°C. At * A curious misstatement of what I said on this point in my earlier paper has been made by Connstein, Hoyer, and Wartenburg (‘Ber. d. d. Chem. Ges.,’ vol. 35 (1902), p. 3988), and recently repeated by Vierling (‘Journ. Suisse de Chim. et Pharm.,’ vol. 42, (1904), p. 391). Iam represented as saying that the action of the lipase is stopped by the liberation of the acids in the endosperm. My paper contains no such statement. What I said was that if the enzyme was set to work én vitro in the presence of dilute hydro- chlorie acid it was rapidly destroyed. Reference to my paper will show that I regarded the organic acids formed in the endosperm helpful and not deleterious —J. R. G. 84 Prof. J. Reynolds Green and Mr. H. Jackson. [Mar. 22, the end of this time the uninjured epidermis ,had produced such a change in the globulin that the solution gave a vivid reaction for tryptophane on addition of a little chlorine water. The contents of the other tube were unchanged. The presence of trypsin in the cotyledonary epidermis was consequently proved. An extract of the cotyledons gave the same results. Taking these experiments in conjunction with Mr. Biffen’s observations, there can be little or no doubt that the special cells alluded to secrete the trypsin. These observations throw a light upon certain phenomena already alluded to, which were first recorded by Van Tieghem,* and subsequently corrobo- rated by one of us.f Van Tieghem dissected the embryos out of seeds of Ricinus and exposed the endosperms on damp moss for some weeks to a temperature of 25 to 30°C. After several days of this exposure he found them growing considerably, and at the end of a month they had doubled their dimensions. The change was caused by the enlargement and partial separation of the constituent cells. In the interior of the cells he found the aleurone grains to be gradually dissolving, and the oily matter to be slowly diminishing. In the confirmatory experiments made by one of us the changes were found to be much more rapid when pieces of the cotyledons were left in contact with the endosperms than when the embryo was entirely removed. No satisfactory explanation of these phenomena was. forthcoming at the time that they were observed, but the discovery that the tryptic enzyme is secreted by the cotyledons affords one. That a very slow germination takes place in the complete absence of the cotyledons may be explained by a small exudation of the enzyme from the latter before their removal or by the endosperm-cells themselves secreting a small quantity of it when the growth of the protoplasm is resumed during the early stages. The diffusion of the trypsin from the cotyledons into the tissue of the endosperm is exactly paralleled by the diffusion of diastase from the scutellum of the barley grain, described by Brown and Morris.t CONCLUSIONS. The germination of the seed of Ricinus is shown by the experiments now recorded to be associated with a remarkable activity of the cells of the endosperm, which spring into renewed life and set up a very complex * Van Tieghem, “Sur la Digestion d’Albumen,” ‘Comptes Rendus,’ vol. 84 (1877), p. 578. + Loe. cit., p. 389. { Brown and Morris, “On the Germination of some of the Graminex,” ‘ Journ. Chem. Soc.,’ vol. 57 (1890), p. 495. 1905.] On the Germination of Seeds of the Castor Ol Plant. 85 metabolism. Their protoplasm grows and takes a prominent part in these metabolic changes, secreting enzymes, and setting up various chemical changes in the cells partly by means of the latter and partly independently of them. In this renewed activity the embryo also takes a share by contributing to the enzyme-formation. The result is the production of a great variety of nutritive material, partly the direct product of enzyme- action, partly produced by the secretory activity of the protoplasm and partly by the interaction of the products of the first two agents. Two varieties of sugar, lecithin, fatty acids, and the products of their oxidation, proteids, and the products of their digestion, including various crystalline nitrogenous bodies, amino- and amido-compounds at least are present. In this mass of nutritive material the embryo is plunged, and by the delicate epidermis of its cotyledons it absorbs, probably selectively, what it needs for its own growth. It is not easy to follow the process of absorption in detail, on account of the metabolism accompanying growth, which is very speedily set up in the cells of the embryo. Analyses of the cotyledons show them to contain a varying quantity of lecithin, amounting in some cases to 1°36 per cent. of their dry weight. Both the sugars can be detected in them, the relative amounts, however, varying, but cane-sugar being usually present in largest quantity. The reaction of the sap is acid, traces of phosphoric acid being mixed with an organic acid whose nature has not been ascertained. In fact, the transport of the nutritive substances to the embryo seems to be much the same in character as their transport in the tissue of the endosperm, Probably in both cases the presence of protoplasmic threads in the various cell-walls plays an important part in the matter; it seems at any rate probable that this agency is necessary to explain the transport of lecithin to the embryo. A very small quantity of lecithin can be dissolved in water or exist as a fine emulsion. It is improbable, however, that it can be trans- mitted through the cell-walls by dialysis alone. Dialysis no doubt plays a large part in the absorptive processes, especially where the crystalline substances are concerned. The renewed metabolism in the endosperm-cells thus furnishes a mass of nutritive material on which both the endosperm-cells and the young embryo feed, and there seems to be no particular difference in the manner in which they are severally nourished. VOL, LXXVII.—B. H 7 ite OP) Ged Ry DA aaa bt pra OH WORE uy Hy pala ; ; y rae eee iw 7 | - teas , a] nf ic ae ¢ | ! te wy ol x..! a tea uA ae Re sli iat, yal ho ees ee ahit cance tg ia} %j Sua : a be ete. 8 ine } bet . tibet 3 me P + Mh ; { a ey i) ‘Xie aii ‘ : stint -depet: We i Die pa ae Pan y ; | ( ' { : 7 r T iG F ' ; 4 ee d abet: Spee Ae BE; t a , ye i sal ms y: ; «nh ab) dee a a HN. i ee DY negra jad it ——— Oo 20 40 60 80 I00 120 140 160 I80 200 220 240 260 280 300 320 340 360 ~<——_Pressuresin mm. of mercury.a——> Fie. 1. Haperiment 2.—Heart Muscle. The hearts of two sheep were perfused with saline, the parts free from visible fat were then cut out, minced and otherwise treated as above described. The percentage of proteid was 6°28 per cent., and of ethereal extractive 1:16 per cent. The results are given in Table II and graphically in Curve 1, fig. 2. 90 Prof. B. Moore and Dr. H. E. Roaf. Table I1.—Temp. 40° C. [Oct. 19, Percentage by Pressure of Percentage by SW chloroform in weight of eines ai vapour space, chloroform originally added, in mm. of pumped off into mercury. vapour space. 0:05 12 62 0 :0078 O71 21°89 0 0135 0-2 47 -02 0 -0290 0'3 75 °35 0 0464 0°5 93 -87 0 0578 0°8 169 °25 0 1042 1:0 198 -49 0 1538 2°0 297 ‘11 0 1829 3°0 322 -94 0 -1988 4:0 339 88 0 -2092 5:0 334 °62 0 -2060 Percentage by Coefficient of weight of distribution chloroform between remaining in vapour space solution. and solvent. 0 0422 1: 5°4 0 0865 1: 6°4 01710 ig 52 0 -2536 abe 159955 0 4422 ie (77 0 6958 aba Ot/ 0 8462 1: 6:2 1°8171 yg (8) 8) 2 8012 1:14:1 38-7908 ato aig} 4, °7940 1 : 23°3 Experiment 3.—Lnver Tissue. Sheep’s liver was prepared in the usual manner. The percentage of proteid was 11:37, and of ethereal extractive 2°80. The results of the experiment are given in Table III, and graphically in Curve 2, fig. 2. Table III.—Temp. 40° C.. Percentage by weight of chloroform originally introduced. ARwWNrOOoCOO SSSSCSHAWNE Pressure of Percentage by Percentage by chloroform in weight of weight of vapour space, chloroform chloroform in mm. of pumped off into remaining in mercury. vapour space. solution. 9-63 0 0059 0 0941 20°85 0 :0128 01872 32 84 0 -0202 0 :2798 56 “41 0 0347 0 4653 85 58 0 0527 0 7473 108 °21 0 -0666 0 9334 161 -26 0 0993 1 -9007 199 -29 01227 2 8773 239 -75 0 1476 3 8524 267 -80 0 1649 4°8351 Coefficient of distribution between vapour space and solvent. :15°9 :15°4 :13°9 :13°4 :14°2 : 14:0 : 19-2 :23°5 : 26-1 : 293 A ee ee pe 1905.] On certain Properties of Solutions of Chloroform, etc. 91 U U Oy) gy / v } solution. ———> 20 40 60 80 100 120 140 160 I80 200 220 240 260 280 300 320 340 360 ~——Pressures in mm. of mercury.—#_ <«—— Percentages of chloroform by weight in Fic. 2. 92 Prof. B. Moore and Dr. H. E. Roaf. [ Oct. 19, B.—Experiments on the Relationship between Vapour-pressure and Concen- tration of Chloroform in Emulsions in Saline of the Ethereal Extractives (Lipoids) of Serum and Brain Tissue. These experiments were devised with the object of testing whether the alterations in the relationships between vapour-pressure of the anesthetic and its concentration, as compared with water and saline, found in the case of serum and the tissues, were due entirely to lipoids or ethereal extractives contained in these fluids, or whether part of the effect was due to action of the proteids. The results shown in the curves of figs. 1 and 3, clearly show that a great deal of the action is due to the proteid. That a certain amount of the anesthetic will be taken up by the lipoid in a physical fashion there can be no doubt on account of the high solubility of chloroform and other anesthetics in such lipoid substances. But we hold that the portion of anesthetic so taken up and held by the lipoid is passive and not active, and that it is the other portion taken up by the proteid (the existence of which figs. 1 and 3 demonstrate) which is active in paralysing protoplasmic activity and producing anesthesia. It is a matter of common experience that the greater the amount of fatty tissue in a subject undergoing anesthetisation, the greater the amount of anesthetic required. The portion of anesthetic which is absorbed by the lipoid is imprisoned as far as purposes of anzsthetisation are concerned, and so much the more anesthetic must be given in order to raise the pressure of anesthetic sufficiently and cause combination between cell-protoplasm and anesthetic with resulting anesthetisation. The ethereal extractives (lipoids) were obtained by the following method :— The proteid of the serum or brain tissue was completely precipitated by addition of excess of absolute alcohol, and the precipitate was separated from the alcohol. The precipitate was thoroughly extracted with ether. The absolute alcohol solution was evaporated to dryness and the residue also thoroughly extracted with ether. The two ethereal extracts were united and the ether evaporated off. The total ethereal extractive was weighed and then made up into a fine emulsion by shaking with normal saline (0°75 per cent.). The volume of the emulsion was made equal to that of the serum originally taken, and in the case of the brain tissue the concentration of the emulsion of the ethereal extract was made equal to the amount of lipoid directly determined in the sheep’s brain. For comparison with the results in the case of the ethereal extractive emulsion of serum, the results in the case of serum from our former paper are given, and the comparison is shown in the two curves of fig. 3. In the case of the brain tissue and the emulsion of the lipoids of brain tissue of equal concentration, the results are shown alongside in the two curves of fig. 1. 1905.] On certain Properties of Solutions of Chloroform, etc. 93 Separation or Coagulation of the Lipoid Emulsions by Chloroform and other Substances. An interesting physical effect is seen as the amount of chloroform added to the emulsion of lipoids in saline is increased. Ata certain stage, dependent upon the richness of the emulsion in lipoid, a complete separation of the lipoid, in a butter-like mass, is obtained, leaving the saline practically free from lipoid. The phenomenon suggests a resemblance to the similar precipi- tation of proteid observed under like conditions, but there is this difference, that in the case of proteid, when the amount of anesthetic is sufficient, the precipitate is permanently altered, being coagulated and rendered insoluble in water or saline, while the lipoid is only physically thrown out, and can be re-dissolved in ether and again made into an emulsion with saline. This phenomenon of physical aggregation of the lipoid by the anesthetic is the more remarkable because the emulsions are exceedingly permanent and remain unaltered for days. The permanency is probably due to lecithin, and the emulsion under the microscope shows small bodies, which are not in most cases spheres, but show the appearance of bi-concave discs of varying size, many being no larger than mammalian blood-corpuscles. The physical cause for the production of such discs is at present unknown to us, but the matter is being further investigated. In the case of the emulsion of brain tissue (containing 4°07 per cent. of lipoid) coagulation or separation of the emulsion occurs in the cold when about 2 per cent. of chloroform has been added, and the coagulation occurs much earlier at body temperature. The coagulum or separated lipoid forms a jelly-like mass, which later separates into a thin whitish fluid and a butter- like mass. These emulsions can also be coagulated or separated by solutions of neutral salts, alcohol, benzol, xylol, and other organic fluids. The precipitation may be due to a lowering of surface tension in the emulsion, and presents an interesting analogy with the precipitation of proteids and other colloids from solution by neutral salts, chloroform, and other organic substances which act as anesthetics. It may be noted in this connection that froth on serum disappears when chloroform is dropped into it, and we have noticed the absence of frothing on stirring on the chloroform side of our densimeter as compared with the control side. Expervment 4.—Ethereal Extract of Serum. The extract was obtained as described, the percentage of ethereal extractive (lipoid) in the emulsion in saline as introduced into the densimeter was 0-206. The results are given in Table IV, and shown graphically in Curve 2, fig. 3, ° where they are contrasted with the results given by entire serum. 94 Prof. B. Moore and Dr. H. E. Roaf. [Oct. 19, Table IV.—Temp. 40° C. Percentage by Pressure of Percentage by Percentage by Coefficient of weight of chloroform in weight of weight of distribution chloroform vapour space, chloroform chloroform between originally in mm. of pumped off into remaining in vapour space introduced. mercury. vapour space. solution. and solvent. 0-1 38 -60 0 -0238 0 -0762 1:3°2 0:2 77°41 0 0476 01524 1:3°2 0:3 108 °32 0 0667 0 2333 1:3°5 0°5 177 “71 0 °1094 0 °3906 1:3°6 0°8 262 ‘60 0°1617 0 6383 1:3°9 1°0 309 :27 0 +1904 0 8096 1:4°3 1:2 351 ‘90 0 °2105 0 '9895 1:4°7 Poel a ae be f / / <— Percentages of chloroform by weight in solution. —> fo} 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 ~«— Pressures.in mm. of mercury. —> Fig. 3. 1905.] On certain Properties of Solutions of Chloroform, etc. 95 Experiment 5.— Ethereal Extract of Sheep’s Brain. The emulsion, prepared as before, contained 4:067 per cent. of ethereal extractives. The results are given in Table V, and are contrasted graphically with the results obtained from the emulsion of the entire brain in fig. 1, Curve 1, brain tissue; Curve 2, ethereal extractives. Table V.cTemp. 40° C. Percentage by Pressure of Percentage by Percentage by Coefficient of weight of chloroform in weight of weight of distribution chloroform vapour space, chloroform chloroform between originally in mm. of pumped off into remaining in vapour space introduced. mercury. vapour space. solution. and solvent. 0-103 32 °44, 0 0199 0 0831 1:4°2 0-206 52°15 0 :0320 0 -1740 1:5°4 0 309 62 12 0 :0381 0 2709 iL faa 0°515 107 ‘11 0 0656 0 4494 1:6°9 0-820 145 °31 0 0890 0 -7310 1:8:2 1-040 190 °22 0°1161 0 9234 1:7°9 The experiment could not be carried beyond this point because the lipoid separated out at 40° C. at all higher concentrations in chloroform. C.—Relative Solubilities of a Series of Anesthetics in Water and Serwm respectively, and the Effects of Addition of such Anesthetics wpon the Serum. The method of determining the solubility of the anesthetic consisted in weighing out known amounts of the anesthetic into water and serum respectively, stoppering the flasks, placing on the mechanical shaker for some hours, and then determining by direct observation that concentration in each case at which the anesthetic ceased to be dissolved. The anesthetics so tested were chloroform, ethyl ether, ethyl acetate, amyl aleohol, amyl acetate, benzol, and xylol. The results obtained were as follows :— Chloroform. Solubility in water at 13° C....... 4 serum _,, 0:95 per cent. 400 ,, In water, 0°8 per cent., all dissolved ; 0-9 per cent., all dissolved ; 1 per cent., not dissolved completely ; estimated solubility, 0°95 per cent. In serum, 3 per cent., all dissolved ; 3°5 per cent., all dissolved ; 4 per cent., all dissolved save a few small globules; estimated solubility, almost 4 per cent. 96 Prof. B. Moore and Dr. H. E. Roaf. [(@et.” 19% Ethyl Ether. Solubility in water at 15° C....... 8 per cent. pA Serum (ene nl eee TM At 10 per cent. and over in the serum there was a slight opalescence, and the fluid began to grow gelatinous ; 11 per cent., completely dissolved ; 12 per cent. not dissolved ; and at higher percentages two phases separated, the layer on top forming a clear, thin jelly. Ethyl Acetate. Solubility in water at 15° C....... 79 per cent. pe SN 5 as baoae 10 At 8 per cent. complete solution of the ethyl acetate and commencing precipitation of the proteid occurred. Solutions of 9 and of 10 per cent. strengths also dissolved completely. Separation into two phases occurred at 12 per cent., the upper layer being viscid. Amyl Alcohol. Solubility in water at 15° C....... 2-4 per cent. 5 Serum), yen) Wes 8 ms or over. At 3 per cent. complete solution of anesthetic with commencing precipita- tion of proteid. This precipitation gradually increased with concentrations of 3°5 and 4 per cent. up to 8 per cent., when a thick cream was formed. A separation into two phases could not be observed at any stage. Amyl Acetate. Solubility in water at 15° C....... 0:25 per cent. e Serum... ut eee 15 Bs At 1 per cent., complete solution of anesthetic with precipitation of proteid ; at 1°5 per cent., all dissolved except a few minute globules. Benzol. Solubility in water at 15° C....... 0-15 per cent. a Seruin > 48) ee 06 a At 0°5 per cent., commencing precipitation and opalescence; 0°6 per cent. completely dissolved; 0-7 per cent., not all dissolved. Xylol. Solubility in water at 15° C....... 0:016 per cent. \. BEL} ih Rye Cee 0:2 i 1905.] On certain Properties of Solutions of Chloroform, etc. 97 At 01 per cent., complete solution, cloudy from proteid precipitation ; 0:2 per cent., all dissolved; 0°3 per cent., not all dissolved. In all cases the solubility in serum is higher than in water, and it is clear from the results that a similar association between the anesthetic and the proteid of the serum occurs throughout the series. The amount of ethereal extractive in the different samples of serum used for the determinations varied between 0:24 and 0°36 per cent., thus showing that the increased solubilities found could not be due to solution of the anesthetic in lipoid or ethereal extractives present in the serum. D. Changes in Depression of the Freezing Point of Water, Saline, and Serum respectively caused by the addition of Chloroform. These experiments were undertaken to elucidate if possible the state in which the chloroform existed in the proteid solutions. On account of the chloroform being volatile, the apparatus had to be completely closed, and the stirring accomplished by electro-magnet in the usual way. The solution was placed in a straight tube, which was almost completely filled, so that no correction was required for escape of chloroform into the air space above the fluid. The changes in depression of freezing point are small, but in every case the results are comparable and concordant. In the table (Table V1) the percentages of chloroform are given by weight, and the figures are corrected to allow for the degree of supercooling. The determinations were made with Beckmann’s apparatus. Table VI.—Changes in A for Amounts of Chloroform added to Various Solutions indicated at the Heads of the separate Columns. Saline g Sor Sonam Distilled water solution Serum Serum cha 2 te eiae plus 0°6 per (0-75 per | plus 0°6 per | plus 1:0 per ot nes an on ap “9 cent. cent.) plus cent. cent. be ey rs As is ae chloroform. oad Reon chloroform. | chloroform. Pi peareeralichionat ae |chicsotecn 0-061 0-081 0-060 0-101 0°135 0-138 0 142 0 ‘075 0-075 0 :062 0 -086 0-141 _— — 0:077 0:072 0:072 0-104 _ = = 0-099 0 082 — — — = = Average 0:078 0078 | 0 ‘065 | 0 ‘097 0188 0°138 0 °142 It is noticeable from the table (1) that the lowering of freezing point for chloroform in serum is less than the lowering due to the same amount of 98 Prof. B. Moore and Dr. H. E. Roaf. [Oet. 195 chloroform in water or saline; (2) that the lowering of freezing point in the case of serum goes on increasing far beyond the point of maximum solubility of chloroform in water or saline, but that the increased lowering is not strictly proportional to the extra amount of chloroform added, but progres- sively less; and (3) above 2 per cent. further addition of chloroform has no more effect upon the freezing point. The latter fact points to a removal of the chloroform by the proteid corresponding to the steep rise in the vapour- pressure curves at|high concentrations. E. Changes in Electrical Conductivity in Saline and Serum produced by the addition of Chloroform. The electrical conductivities were determined by the method of Kohlrausch, at the temperatures of 0°, 15°, and 40°C. In the case of saline (0°75 per cent.) and different samples of serum, both before and after addition of definite weighed amounts of chloroform. In the case of the saline the amount of chloroform added was 0°6 per cent. by weight, and in the case of serum amounts of 0°6, 1, 2, and 3 per cent. by weight were added in succes- sive experiments. On account of variations in the different samples of serum there are considerable fluctuations in the results, but certain facts are observable. In Table VII the alterations in the value of K (the specific conductivity, multiplied by 10”) are shown, which result from the addition of the amounts by weight of chloroform shown at the head of each column. It is noticeable that the conductivity is reduced without exception in all cases where chloroform is added to normal saline, due to a diminution of the ionisation of the sodium chloride by the chloroform and alteration of ionic velocity. But when chloroform is added to serum there is, in many cases, an actual increase in conductivity, and in other cases the diminution is less than occurs with the corresponding amount of chloroform added to normal saline. This indicates that there is a tendency for inorganic salts to be set free from the proteid and add to the conductivity. This is only seen as an actual increase where it outbalances the action above mentioned of the chloroform in reducing conductivity in pure saline solution. The table shows that the amount of increase in conductivity varies with the temperature and the amount of chloroform added. Thus at 0° C. the diminution in conductivity is the same in saline and serum for 0°6 per cent. of chloroform, but when 1 per cent. of chloroform has been added, the diminution becomes converted into an increase; again, at higher percentages, a diminution is seen, as if all the electrolyte possible had been detached from the proteid at the lower concentrations, and the diminution was now that due to the effect of the additional chloroform upon the saline in solution (fig. 4). 1905.| On certain Properties of Solutions of Chloroform, etc. 99 Table VII.—Changes in K x 10? for amounts of Chloroform added to Serum and Saline. Saline solution (0°75 per cent.) | Serum plus Serum plus Serum plus Serum plus Temperature. plus 0°6 0°6 per cent. | 1:0 percent. | 2:0 percent. | 3:0 per cent. per cent. chloroform. chloroform. chloroform. chloroform. chloroform. © Oh (0) — 0:0140 — 0:0267 + 0:0015 — 0:0275 — 0:0176 0 — 0:0204 — 0:0077 + 0:°0160 — 0:0139 — 0:0275 Average .... — 0°0172 | — 0:0172 | + 0:0087 ge 0 0207 — 0:0225 15 — 0:0360 + 0:0390 + 0:0440 + 0:0140 — 0:0060 15 — 00020 + 0:0460 + 0:0890 — 0:0050 — 0:0130 15 — 00361 + 0-0032 — 0:0146 — 0:0255 — 0:0336 15 — 0:0200 — 0-0092 — 0:0044 — 0°0162 — 0:0323 Average . — 0:0235 | + 0:0197 + 0:0285 | — 0:0082 | — 0:0212 40 — 0:0270 — 0:0240 — 0:0320 — 0:0410 == 40 — 0-0300 — 0:0270 — 0-0880 — 0:0690 ae 40 — 0:0410 — 0:0140 — 0:0470 — 0:0650 — 40 — 00399 — 0:0170 — 0:0331 — 0:0577 — 0:0674 40 — 0:0314 — 0:0061 — 0:0312 — 0:0669 — 0:0488 Average ‘ — 00339 — 0:0176 | — 00463 | — 0:0599 | — 0:0581 OaG i O-o1 OD bret rve cle ate 2 sd O05 3:0 Percentages of ehigroronm by weigce in solution. Change in specific conductivity x 10%. Fic. 4.—Changes in Electrical Conductivity on addition of Chloroform to Saline (Curve 1) and Serum (Curve 2) at 0° C. Between the first two points the curves are identical. The effects are most marked at 15° C.; here, instead of the diminution occur- ring on addition of chloroform to pure saline, there is found an increase in conductivity in the case of serum both for addition of 0-6 per cent. and 1 per 100 Prof. B. Moore and Dr. H. E. Roaf. [Oct. 19, cent., which as before becomes converted into a diminution for 2 per cent. and 3 per cent. (fig. 5). At 40° C. there is no actual increase at any concen- Q fe) =) 5 Chee in specific conductivity x 107. fo) fe) nN 0°03 Rorooe ee of aoe by oat in Sanne Fie. 5.—Changes in Electrical Conductivity on addition of Chloroform to Saline (Curve 1) and Serum (Curve 2) at 18° C. @ ° - 2 fe) iy Change in specific conductivity x 10*. ° c} ur 0:06 O5 TO 2:0 30 Percentages of chloroform ly weight in solution. Fic. 6.—Changes in Electrical Conductivity on addition of Chloroform to Saline (Curve 1) and Serum (Curve 2) at 40° C. tration, but the fall in conductivity is less for serum than saline at 0°6 per cent., and this fall is increased at the higher concentrations in chloroform (fig. 6). 1905.] On certain Properties of Solutions of Chloroform, etc. 101 Taking the experiments as a whole, there is evidence of two opposing factors partially masking each other, the first being due to a setting free of electrolyte from the proteid by the added chloroform increasing the con- ductivity, and the second the well-known action of a non-electrolyte in diminishing the conductivity, which is seen clearly in the case of the pure saline plus chloroform. Summary and Conclusions. The experiments recorded in the present communication support the conclusion drawn in our previous paper that anesthetics form unstable compounds or aggregates with the proteids of the tissue cells, and that anesthesia is due to a paralysis of the chemical activities of the protoplasm as a result of the formation of such aggregations. The comparative experiments with ethereal extracts demonstrate that the action is upon the cell proteids and not upon the lipoids. The compounds or aggregations so formed are unstable, and remained formed only so long as the pressure of the anesthetic in the blood is maintained. The results of our experiments may be summarised as follows :— (1) The solubility of all anesthetics experimented with is higher in serum than in water. (2) At a certain concentration, definite for each anesthetic, there occurs opalescence and commencing precipitation of proteid. (3) At equal concentration of chloroform in water or saline on the one hand, and serum, hemoglobin, or the tissues (brain, heart, muscle, and liver) on the other, the vapour-pressure is always higher in the former than in the latter. (4) The curve connecting vapour-pressure and concentration is, in the case of water and saline, a straight line; while in the case of serum, hemoglobin. and the tissue proteids it is a curve showing association, especially at the higher concentrations. (5) Comparative determinations of vapour-pressure and concentration, in serum and brain tissue and in ethereal extracts of these equal in concentration of lipoid, show that the proteid of the tissue combines with the anesthetic. (6) Determinations of the effects of addition of chloroform upon the lower- ing of freezing point confirm the results obtained by the vapour-pressure and solubility determinations. (7) Determinations of the changes in electrical conductivity caused by addition of chloroform indicate that accompanying the combination of the anesthetic with the proteid there takes place a splitting off of electrolytes. VOL. LXXVIIL—B. I 102 Prof. Moore, Dr. Roaf, and Mr. Whitley. [ Oct. 9, (8) When the lipoids, extracted from serum or tissues by ether are made up into an emulsion with normal saline, many of the lipoids take the form of bi-concave discs. (9) The lipoid emulsions are very permanent, but separate on the addition of anesthetics or neutral salts, in similar fashion to colloidal solutions. On the Effects of Alkalies and Acids, and of Alkaline and Acid Salts, upon Growth and Cell Division in the Fertilized Eggs of Kchinus esculentus.—A Study in Relationship to the Causation of Malignant Disease. By Benyamin Moore, M.A., D.Sc., Johnston Professor of Bio-Chemistry, University of Liverpool; Herbert E. Roar, M.D. Toronto, Johnston Colonial Fellow, University of Liverpool; and Epwarp WuirtLey, M.A. Oxon. (Communicated by Professor W. A. Herdman, F.R.S. Received October 9,— Read November 23, 1905.) The results of observations previously made in the Bio-chemical Laboratory of the University of Liverpool have shown that free hydrochloric acid is absent from the gastric contents, or greatly reduced, in nearly all cases of malignant disease, no matter where the malignant growth happens to be situated. In the paper describing these observations it was pointed out that the most probable cause of this absence of the free hydrochloric acid was an increased alkalinity of the blood-plasma, as a result of which the hydrogen ion concentration in the plasma was so far reduced that the oxyntic cells were no longer able to separate an acid secretion from it.* It seemed to us, therefore, desirable to test the effects of alterations in the concentration of hydrogen and hydroxyl ions respectively, upon the growth and cell division in some organism in which cell division was proceeding rapidly, and which could be easily subjected directly to changes in acidity or alkalinity of the medium in which it was living. We selected for this purpose the fertilized eggs of a species of sea-urchin (Echinus esculentus), because at the season of the year at which our experi- ments were carried out (April, 1905) the ripe eggs can readily be obtained from the gonads of the female, and be fertilized by mixing with the sperm similarly obtained from a ripe male. Hence from the single cell stage * “Roy. Soc. Proc.,’ B, vol. 76, p. 138, 1905. 1905.] Liffects of Alkalies, etc., on Eggs of Echinus. 103 onward the rate of growth and any irregularities in cell division can be observed, and the effects of the addition of alkali or acid, or alkaline or acid salt, can be compared under exactly similar conditions, and contrasted with a control grown alongside in unaltered sea-water. The experiments were made, by kind permission of Professor W. A Herdman, F.R.S., at the Marine Biological Station, Port Erin, Isle of Man, during the month of April, 1905. J. Loeb,* in 1898, reported experiments on the influence of acids and alkalies on the development of the larve of sea-urchins, showing that the addition of acids to the sea-water delayed the development, and finally, at a certain concentration of hydrogen ions, inhibited the development com- pletely. With the addition of alkalies the development during the first 12 hours was not at all hastened, or hastened to a hardly appreciable extent. On the second day, however, and sometimes also even on the third day, the eggs that developed in sea-water which had been made alkaline were occasionally, but not always, in advance of eggs of the same brood which had been raised in normal sea-water. At the time of the publication of this paper, Loeb held the current view that sea-water had an alkaline reaction, and accordingly drew the conclusion that, for developmental processes, it is necessary to have an alkaline reaction or, in other words, a higher concentration of hydroxyl ions than exists in distilled water. At a later date Loeb,f however, came to the conclusion that normal sea- water must be regarded as a practically neutral solution, and hence ascribed the favouring action of alkalies upon cell-division and growth to the neutralizing action of the alkali added upon acid products formed by the cells in the process of growth which would otherwise cause an accumulation of hydrogen ions and arrest the development. This opinion is based by Loeb upon the results of the observations of Friedenthal,t Fraenkel§ and Farkasj| which indicate that blood-plasma possesses no higher concentration in hydroxyl ions than distilled water ; upon observations made by Dr. Cottrell at Loeb’s suggestion upon the sea-water of the Bay of San Francisco by means of hydrogen electrodes, which showed * © Arch, f. Entwickelungsmechanik d. Organismen,’ 1898, vol. 7, p. 631. + ‘Arch. f. d. ges. Physiol.,’ 1903, vol. 99, p. 637; zb¢d., 1904, vol. 101, p. 340; zbed., 1904, vol. 108, p. 506 ; ‘Univ. of California Publications, Physiology,’ 1908—4, vol. 1, pp. 39, 139. t ‘Arch. f. Anat. u. Physiol., Physiol. Abth.,’ 1903, p. 550. § ‘Arch. f. d. ges. Physiol.,’ vol. 96, 1903, p. 601. || ‘Arch. f. d. ges. Physiol.,’ vol. 98, 1903, p. 551; see also Hiber, ‘Arch. f. d. ges. Physiol.,’ vol. 99, 1903, p. 572. it 7 104 Prof. Moore, Dr. Roaf, and Mr. Whitley. [Oct. 9, that the concentration of hydrogen ions was greater by one power of ten than in distilled water ; and upon the observations of Loeb himself, (1) that normal sea-water is practically neutral to phenol-phthaléin, and (2) that addition of sodium bicarbonate or of di-sodic phosphate, which from their neutral reaction to phenol-phthaléin are regarded by Loeb as neutral salts, produced the same favouring effect upon cell-division and reproduction as the caustic alkalies. A solution of the mixed phosphates or carbonates in which there is an approximate balance between the concentration of hydrogen and hydroxyl ions such that these concentrations are nearly equal, cannot, however, be regarded as neutral in the same sense as distilled water is neutral; or as being acid or alkaline in the same sense as a solution containing only free acid or free alkali can be regarded as being acid or alkaline. Nor will such a solution of phosphates and carbonates as is present in blood- plasma or sea-water have a similar action upon living cells to either distilled water or a neutral solution of such salts as sodium chloride of equal osmotic concentration. Such solutions form a peculiar type of their own, which possess the property of behaving either as an acid or as a base. Such solutions can behave in this manner to any substances possessing weak acid or basic properties brought into the same solution, and it is for this reason that such solutions behave entirely differently to distilled water in their associating or dissociating action upon coloured indicators, and so indicate at the same time an acid reaction to one indicator and an alkaline to another. A quantity of acid or of alkali can be added to the solution, which would cause an enormous variation in the relative concentration of hydrogen and hydroxyl ions if added to distilled water, without altering to anything like the same extent the relative concentrations of these two ions. Therefore blood-plasma, and to a less extent sea-water, possess, on account of the mixed phosphates and carbonates which they contain, a steadying action upon variations in the concentrations of the hydrogen and hydroxyl ions. When acid or alkali is added to the plasma, instead of there occurring a corresponding swing in the concentration of the hydrogen and hydroxy] ions, there takes place an alteration in the equilibrium of the ions of the phosphates and carbonates, which neutralizes, in great part, the hydrogen or hydroxyl ions added, and prevents the plasma becoming markedly acid or alkaline. Without such a controlling action the life of the cells would be rendered impossible, for, as our experiments show, the living cell is most sensitive to even small variations in either hydrogen or hydroxy] ion. This powerful action of alterations in concentration of hydrogen or hydroxyl 1905. | Liffects of Alkalies, Go, on Lggs of Echinus. 105 ion, arises from the fact that the constituents of the protoplasm behave like very weak acids or bases, and are affected by variations in hydrogen or hydroxyl ions in a similar manner to coloured indicators. Very small variations in ionic concentrations compared to those found in even dilute solutions of free acid or alkali, will hence cause the constituents of the cell to become almost completely associated or dissociated, and prevent the chemical reactions from occurring which are necessary for the metabolism and life of the cell. This is shown in our experiments by the very small amounts of either acid or alkali which suffice to kill the cell. Within the limits at which life is still possible, but at which the concentrations of the two ions are varied from the normal, profound alterations occur both in the rate of growth, and in the type of the cell divisions. In the case of malignant growths similar variations in hydrogen and hydroxyl ions occur as shown by the absence of the acid in the gastric juice. It is from this point of view that we have studied the effects upon growth and cell-division of the addition of acids or alkalies in small amounts to the medium in which such processes are taking place. The problem before us is that of the effects upon the cell of variations in the hydrogen and hydroxyl ion concentration, and for the purposes of such an enquiry it is not neccessary to know what are the actual concentrations of the two ions in normal plasma, nor to discuss which of these two magnitudes is the greater. If alkali be added to plasma or to sea-water the concentration in hydrogen ions will fall and that in hydroxyl ions will rise, and conversely on adding acid a reverse change will take place. The experiments we have made show the results of such changes upon cell growth and nuclear division. A number of other observers have drawn attention to irregularities in cell growth and nuclear division induced by the action of foreign chemical substances upon the dividing cells. Thus, O. and R. Hertwig,* and also Galleottit have shown that pathological mitoses with irregular division in number and amount of the chromosomes may be induced by such substances as quinine, chloral, nicotine, anti-pyrine, cocaine, and potassium iodide. The derangements in division so produced closely resemble the pathological divisions described in cancer cells by Klebs,t Hansemann,§ and Galleotti| * *Jenaische Zeitsch. f. Naturwissenschaft,’ vol. 20, 1887, pp. 120, 477. + ‘Beit. z. path. Anat. u. z. allgem. Path.,’ vol. 14, 1893, p. 288. t ‘ Die allgem. Pathologie,’ Fischer, Jena, 1899, Part 2, pp. 528, et seg. § ‘ Virchow’s Archiv,’ vol. 119, 1890, p. 299 ; vol. 123, 1891, p. 356. || Loe. cit. 106 Prof. Moore, Dr. Roaf, and Mr. Whitley. [Oct. 9, Hansemann recognizes in such abnormal mitoses two general groups, (1) asymmetrical mitoses, in which the chromosomes are unequally ditributed to the daughter cells, and (2) multipolar mitoses, in which the number of centrosomes is more than two, and more than one spindle is formed. Hansemann and Galleotti find in this asymmetry of mitoses an explanation of the well-known fact that in cancer cells many of the nuclei are especially rich in chromatin (hyperchromatic cells) while others are abnormally poor (hypochromatic cells). According to Galleotti, the asymmetrical mitoses which may be artificially produced in the epithelial cells of salamanders by treatment with dilute solutions of anti-pyrine, cocaine or quinine are exactly like those seen in carcinoma. In these experiments, the drugs used for inducing pathological mitoses are not such as can occur in carcinoma. A variation of the concentration in hydregen and hydroxyl ions can, however, occur in the plasma, and, in fact, the observations on the absence of acid in the gastric secretion in cases of - carcinoma make it probable that such variations do occur. It is hence of interest that in our experiments detailed below, in which the concentration of hydrogen and hydroxyl ions was artificially varied, we have found that with increased alkalinity, at a point just short of that at which cell growth was stopped, such pathological mitoses do occur. We have observed both the asymmetrical mitoses with unequal distribution of chromosomes and only two centrosomes, and the multipolar mitoses with three or more centrosomes. Also, as the alkali was increased above the normal of sea-water, a marked tendency to irregularity in size and shape of the resulting cells was observed similar to that seen in cancer cells. As the amount of alkali was increased, there occurred also a shortening of the dividing chromatin rods, similar to that seen in most maturation divisions, until the rods became in some cases converted into rounded dots. In a certain percentage of the divisions, the number of chromosomes was reduced. The number of chromosomes is exceedingly difficult to count with certainty, but the reduction in many cases amounted approximately to one-half the normal, EHaperimental Methods. We have investigated in our experiments the effects of the following alkalies and acids, and alkaline and acid salts :— Sodium hydrate, potassium hydrate, calcium hydrate, ammonia, hydro- chloric acid, acetic acid, carbonic acid, sodium carbonate, sodium bicarbonate mono-sodium phosphate (NaH,PO,), and di-sodium phosphate (Na,HPO,). The stock solutions of the acids and caustic alkalies were made in distilled water, titrated and standardized to normal strength, and from these deci- 1905. | Effects of Alkalies, etc., on Eggs of Echinus. 107 normal solutions in distilled water were prepared. The carbonates and mono- ' sodie phosphate were made up in decimolecular strength, and the di-sodic phosphate in 54, molecular strength, on account of its lower solubility. A saturated solution of calcium hydrate was prepared in distilled water decanted from the lime, and standardized against decinormal acid, and this standardized solution was used in small measured amounts for addition to the sea-water containing the samples of eggs as in the other cases. Similarly, in the case of the carbonic acid, a comparatively strong solution of carbonic acid was prepared by passing the gas through sea-water ; this solution was standardized at once with decinormal alkali, using phenol-phthaléin as indicator, and immediately added in appropriate small measured quantities to the various measured sea-water and egg-mixtures. It may be pointed out that there was no appreciable variations in osmotic pressure of the sea-water caused by the addition of the reagent solutions, because the volume added was small in comparison with the volume of sea- water and egg-mixture to which it was added; further the decinormal solu- tions added, though hypotonic, do not lie very far below the molecular concentration of the sea-water. If the decinormal solutions had been made in sea-water instead of distilled water, an equal or greater amount of change in osmotic pressure would have resulted on account of their being hypertonic, and also precipitation of constituents of the sea-water would in certain cases have occurred. The change in osmotic pressure was, however, in all cases quite a negligible quantity. The experiments involved the use of a large number of vessels on account of the long series of mixtures of sea-water and fertilized eggs with their varying amounts of added chemicals of different kinds, and we had no* anticipated this heavy demand when starting upon our expedition. Hence, we had to make use of such materials as we could find in the Marine Biological Station, or obtain at Port Erin. The earlier experiments were made in well washed out glass jam-pots such as are used for collecting fresh marine specimens, but later we found ordinary plain tumblers of the usual size most convenient. The amount of surface for aeration compared to the volume of fluid was the same in each case throughout each series of experiments, and henee the results obtained are strictly comparable with one another. The method of procedure in starting an experiment was as follows :—The shells of a number of Echini were cut open circularly so as to expose the uninjured gonads, until a ripe male and a ripe female had been obtained as shown by examining under the microscope, and ascertaining that the spermatozoa were active, and the eggs of mature size and well formed. 108 Prot. Moore, Dr. Roaf, and Mr. Whitley. — [Oct 9, Usually two gonads from the female, and about half to one gonad from the male were taken, each in a separate tumbler, gently rubbed up with a small quantity of sea-water, and then separated from débris by filtering through a coarse piece of muslin. The eggs were somewhat diluted with sea-water, a quantity of sperm added, and a drop of the mixture taken out and examined with the microscope until it was ascertained that the eges had developed their fertilization membranes. A number of tumblers corresponding to the number of controls and the total of the various dilutions of the different chemicals to be tested for their effects upon the growth of the eggs had previously been arranged and numbered. The mixture of fertilized eggs was now diluted to a larger volume with sea-water so as to afford 200 cc. for each tumbler, and this volume was measured out into each of the tumblers. As rapidly as possible the desired amount of each chemical was added in each case, and the time of starting noted. The progress of development was observed and noted a few hours after starting the experiment, again the following morning, and so on. In a few cases instead of diluting the egg-mixture after fertilization to a large volume, the proper amount of sea-water was measured out into each tumbler, then an equal volume of the egg-mixture was added to each, followed by the desired amount of the chemical solution. For the purpose of examining the progress of development a dip was taken out by means of a small pipette into a, watch-glass and examined under a low power of the microscope. The state of development of the growing embryo was noted especially with regard to relative rate of growth in presence of the various strengths of the different alkaline and acid solutions. The number of cells was counted in the earlier stages, or the number of cells in an optical section of the circumference of the blastule in the later stages; the commencement of ciliary motion was noted, and the stages in the development of the gastula in those cases where the larve developed so far. Also any irregularities in shape and size of cells in the different cases were noted. In certain cases, the progress of any change in chemical reaction was noted by adding indicators in parallel experiments carried out alongside. Interesting results as to the action of the indicators themselves were so obtained. It was found that the reaction to “di-methyl” did not change throughout the experiment, but the reaction to phenol-phthaléin, which was faintly alkaline even in normal sea-water at the commencement of the experiment, slowly changed towards the acid side. As it was found, however, that phenol-phthaléin even in very small amounts inhibited and caused irregularities in cell-division, the method was adopted of testing quantita- 1905. | Effects of Alkales, etc., on Eggs of Echinus. 109 tively by titration the reaction to both phenol-phthaléin and “ di-methyl” at the end of each experiment after growth had stopped. It was found that even where alkali had been added at the commencement of the experiment the reaction to phenol-phthaléin had become slightly acid at the end in most experiments. Since the amount of alkalinity to “di-methyl” had not altered throughout the experiment, but gave in all cases very approximately the alkalinity of sea-water plus or minus the alkali or acid artificially added, it follows that the change in reaction to phenol-phthaléin must be due to a very weak acid given off during the development of the embryos. Since carbon-dioxide produces exactly the same effects, and respiration must occur in the process of development, it is almost certain that the change in reaction must be due to production of carbonic acid. On account of this natural production of carbonic acid, the results of the first twenty-four hours’ growth are the most valuable, because here the alkaline reaction to phenol-phthaléin persisted, and the production of carbonic acid was small. After preliminary experiments had determined the range at which the developing embryos were definitely affected by the different chemicals, a few final experiments were made in which only two or three concentrations of each chemical were included lying at about the proper range, and the last such experiment was interrupted when the more advanced sets of embryos had reached the morula or early blastula stage.* The embryos were fixed in Flemming’s and Hermann’s fluids, embedded and cut in paraffin, and stained for nuclear division by Heidenhain’s iron-alum and hematoxylin staining, following Flemming’s description. The drawings illustrating the paper were made from this series of preparations, with the exception of fig. 18, which was drawn from the fresh growing cells. Reaction of Sea-Water of Port Erin Bay—One hundred cubic centimetres of sea-water was taken and titrated against decinormal caustic soda solution, (a) with di- methyl- amido- azo- benzol as indicator, and (0) with phenol- phthaléin as indicator. With the “di-methyl” in a first trial 2°35 c.c. were required, in a second 2°37, therefore the alkalinity was = 0:00236 normal. With the phenol-phthaléin 0:24 c.c. and 0:22 c.c. were required, alkalinity = 0°00023 normal. * After cutting sections, it was found that many of the cells described in Table XI (pp. 123, 124) in the fresh condition as morule really possessed in section a small central cavity, and so are termed blastule in the descriptions of the drawings. The organisms described as blastule in the tables showed in the fresh condition an outer layer of cells. 110 Prof. Moore, Dr. Roaf, and Mr. Whitley. [Oet. 9, EXPERIMENTS WITH Table IL— Experiment No. 1, | : ate | Amount of sodium Amount of added alkali Observations of condition No hydrate in decinormal in the solution, ex- | ; solution, added to pressed as a fraction of 200 c.c. of sea-water. normal strength. 4 hrs. 17 h. 50 m. | 1 Control — Four-celled stage | Blastule ............ | 2 0-2 c.c. 0 -0001 Four-celled stage | Blastule ............ 3 | 0-4 ,, 0 0002 Four-celled stage | Blastule ............ 4 | OSes 0 :0004 Four-, six-, and| Blastule. Further eight - celled| advanced in cell- stages. Irregu-| division than con- lar divisions trol 5 1‘4 ,, 0 :0007 Four-celled stage.) Blastule. Further | Irregular divi-| developed than | sion of cells control 6 2°0 ,, 0 0010 Majority two-| Blastule. Further celled. A few} developed than four-celled. control * In order to distinguish throughout between different sets of eggs, each lot of fertilized It is to be noted in this experiment that there is a distinct favourimg action of the added stages of No. 6, but not in sufficient amount to stop growth (see Experiment 2), and as the alkali Also Nos. 4, 5 and 6 remain alive and develop further than Nos. 1, 2 and 3. No. 5 was, at the end of the period 91 h. 50 m., divided into two portions of 100 c.c. each, it. The subsequent development of 50, which had the alkali added, was much more rapid than 1905. | Effects of Alkalies, etc., on Eggs of Echinus. 111 Caustic ALKALIES. Sodium Hydrate. Brood* of Eggs No. 1. of embryos, at interval after start of experiment given at head of each column. 24h.50m.| 29h.15m. | 41h. 45m. | 70h. 30m, 91h. 50 m. 114 h. Blastule. A| Blastule. All| Blastule. All| Commencing gas-} All dead, and — few moving| in active | in active | trule. Nearly| degenerating motion motion all dead Blastulez. A} Blastule. All| Blastule. All | Commencing gas-| All dead, and _— few moving| in active | in active | trule. Nearly) degenerating motion motion all dead Blastule. A/| Blastule. All} Blastule. All | Commencing gas- — — few moving | in active | in active | trule. All dead motion motion Blastule. A} Blastule. All| Blastule. All| Well -developed| Late gastrule. | — few moving} in active | in active | gastrule. All| Very active motion motion active (glass broken by accident) Blastule. Blastule. All| Blastule. All} Well - developed} Late gastrule.| Forming Motionless | in active | in active | gastrule. All| Not so far de-| plutei motion motion active veloped as No. 4 Blastule. Blastule. All| Blastule and | Well - developed} Late gastrule. — Motionless | in active | some gastru-| gastrule. All| (Formalin motion le. All in| active added by mis- active motion | take) eggs is numbered as belonging to the same brood. alkali in the earlier stages, especially observable in No. 4; the alkali is in excess in the earlier becomes diminished by the CO, given off in growth, No. 6 increases and passes the control. of which the first (5a) was left unaltered, while the second had 0°5 c.c. of N/10 NaOH added to that of 5a, and the embryos remained alive much longer. | Oct. 9, Prof. Moore, Dr. Roaf, and Mr. Whitley. 112 ‘SNOIAGO ST SL9T]4O 944 PUB [O1}TOD OT] MEOMJeq 4sBIZUOD 9179 Inq ‘T qUOTMIedxy UI ULT AeT[T¥a SelIEsS SIYY JO [01]UOD oY4 UL pue UB OF oUIvD yUoMdOTeEAGG guourdojoa -op ToqyINT ON quoudojea -op 1eqqInz ON xenyseyq ou ‘ BTN.AOUL MOT WV peop fapnqselg, “UW OF “TY FZ dnoas pe][eo-xIs ouo pus ‘omy OFUI paprarp Aqeyazed ono ydeoxe ‘stiqep snout, SSRUL -B[od UL Ss[[ed poyvpNoed o[suIs [[y | SnoUNyLjeS v Ur s][e0 o[SUIS [TV 9Z00- 0 “ @ , " weas dnoxd pe][90-90179 eo pue ezis ut Tenboun «190A po][e0-0M9 OU “UOIVMOTEr we—nsex | sy[e0 OM] OY YIM ‘Z “ONT UL SUIMOYS souvIsqns ][99 ‘sIqep | Ue, paTjed-omy etouL yeqMoUTOS SuIyOo|-snoulyejes B Ul ‘sT[e0 eTsuIg|4nq {pejeo-ejsuIs [[e Ajay 0Z00: 0 “ep g gynsdeo ut sT[e0 pe][e0-«noF wemsSoart [euors pezis-Aypenboun ssey 10 uezop @ dur | -8000 put ‘pal[eo-om4 BIoAes -U1BJU0D SIOYJO fB[NIOUL OMY IO oUGQ | yNq ‘peTfed-eTsuIs [Te AjxvoyT ST00- 0 00 g 4 (mooes [woydo Jo soUeTe;UMOAIO UT [29 LT) epoyseiq Apwwo pue apaopy [ts 99B48 poTjed-omLy, _ JorqyUOD I ‘Ul OF “TY LT ‘sq G “qg8u0e14s 5 a ; [eurL0t Jo WorRsy 1oyBM-COs JO 70° 0OZ ® st possordxa BAPE OS *I9qUIN NT *qUOUIIIEdKe JO FUOUMLADUOUIULOD 10978 SUIT] pozugs ye soAIQUIE JO SUOIYBALESGO “aorqnyos 044 Ul TTeyIs poppe jo yunouy [eULLoutoep ur eyevrpéq uIMNIpos Jo aNoULY “MOT ATOATSSOOXO T]IJS SI TTeH[e JO JUNoMe oYy yey poyou oq 0} SI 41 4nq ‘peppe oIOA T[eye JO syuNowe Josie, quourredxo styj ul ‘T yuouIedxqy Ul poppe useq you pey yueutdopeasp doqs 03 Aressooou ayvipAY UINIpPOS JO JUNOML WINUIXeU 944 4VY} UMOYS SuUIAvY JUoUTIedxe snotacid oYT, “Z ON poorg ‘Z ‘oN quouwttodxq ‘II 198 113 Liffects of Alkahes, etc., on Eggs of Echinus. 1905.] ose e[nzout ATIve pessed OABY OUONT “SUTZRI eTn.10uL Solos OT} jo wey Xq pooueape ssoy ‘apoio A719 Axes pus ‘aTqeyunoo Jepureuoy “ATWO e3e 4s — -euedep pue pra | Ajxee pure ‘poyjeo-omy, | poyjeo- omy UL «FBT STOO: 0 Sutouemur09 Ayarvo AoA Aue UL AyWO —- woaoyUOYOTB eseqs vpnaysed Sutaou | ‘sesujs yuotegip 4e qm “emayses Ajaeg | A4jrwo ur Agrofeyy | qgok gou apnyserg | emroyy -apnysetq oN OL00- 0 TWLLOF 07 SUTOMEULULOD Jouyuo ueyy SULIe YIM ‘apnty | orJWOD UBYyy pooweA JUOTIOAOUE GATOR | pooUBAPS oAOTY “BlNY “sv poouvape oloy[ | -pe eto ‘@pnaysey | Mtoa ul @pqsey | -se[q pedojesep-[12 MA G000- O poeye,duroo UOLEQUOTLY “BpA14 Z ON Ueyy poouva queuleAout ery -svS podopeaop-]Jo4 | -pe ssey ‘xfmaysey | oaryoe ul apnaqseg | -svtq poedoyeaep-[[o MA _ ‘sa 76 “UL O§ “Y 99 “wW SoP UT LP “uw oF “4 6T "yySueqys “quouti1edxe JO JUOTMODUOUIULOD 104FB SOUITZ pozeqs 4v sofqute JO SUOTIVALOSGO [eULi1ou Jo uoMoRrT ®@ sv possordxe “uorynyos ey Ur TTeyTe peppe jo yunowmy 09 T Jorym0RD *109BM-BOS JO ‘9°D YOST 04 poppe ‘uor4ynTos [ewsouep ut oqv.tpéq UINIpos Jo yUNOW WY ‘) ‘ON poorg ‘¢ ‘ony JUeUMIIEdxa] ‘TIT PCL [Oct. 9, Prof. Moore, Dr. Roaf, and Mr. Whitley. 114 quourdopea ‘ep 0yyMZ; ON preg “ny “ses sutomeurutoD preg ‘xp “ses duToMeuLULOD aniseed AT0 euLos pus ‘aapnqselg peog 9 ‘“spnysvlg 9 ‘ON 8B oureg qnoqe SUIYLOH SslIqep JO sossvu Sutyoo] -snouyejes pues ‘dn woxo1q TLV qOGB SUTULUMIMS ST[90 Very pus ‘dn uoeyorq Aue ‘poreo -100}X18 0} dn suorsIArp [[e0 pus sdnoas xvjnseaat L1eA pus ‘porjeo -OM} IBNSOL MOT YW ‘apAoUM ON pedopeaop moz gnq £ enyselq Seltes JO podojeaep qsogy “(e0UexXeJUMOIIO Jo UOTYe8 [woydo ur sq[eo 9%) a#Tnysv[q o4uT JotyUOo ueYyy peouvape exo, ‘epngseyq ATreRy (Ch chchintiania) Jo wuonoos jeoydo ur s[[oo gT 0% 41) Brnysvyq ATrve L104 pus aco py “WH OF “UY FZ uw Og Y LT peprAtp euou ‘aTqIstA s]jeo moy AOA 20.1} OFUI WOISTATP JO ssooord Ul eUO ‘sT]e0 omy orR o.L0TT3 etey a ‘esva ouo ur 4deoxo o[Suts ere osoyy faTqISTA sT[oo Moz Ato A UOISTATP Lepnsortt Aq sosvo MoT ® ur porusdutoooe ‘dn Suryverq s]]}eo pus aefnuRIs [90 Jo 8{U9JU0L) “peT[ed-cNoF 10 -oMy MOF Y “poTjeo-oysuts [B ATreo yy PeT[99 -MOF IO -9e144 oUIOS “poTjeo-oMT, poT]20-moz IO -901]} [BUOISB000 ‘poT[e0-omZ, poT[eo-«n0F IO -90144 [BUOISvI00 ‘peT[e0-omMG, sereeeseseeecessssees QOBag POT]90-OAT, “Ul O€ “TP ‘qUOUMIIEdKO JO FUOULOOMOUMULOD 10478 SOUITY poqeys 4B soATQUIA Jo SMOTYBAMESGO $200: 0 0600-0 STO0- 0 OTOO- 0 ¢Z000- 0 000.0 "yy due148 [eUIou Jo uoMOBAT ® 88 possordxo ‘HOIyNTOS oY4 UL TVH]S peppe Jo yunoury ‘GZ ON poorg ‘T quows0dxy ‘oyerpAH winisseyog— AT 2Ge, “ G “cc p “ce ¢ ce z 6e GT Jouju0D “TOYBM “808 JO ‘0°) QOZ 09 pepps ‘ao4n{os [eutz10u “ep ur oyeaphy umissejod Jo yunouLy “tequan jy 1905. | Effects of Alkalies, etc., on Eggs of Echinus. 115 Table VimAmmonium Hydrate. Experiment 1, Brood No. 3. | i t of | Amount of f : ;| Amount or | added alkali | Observations of embryos at stated times after commencement SE LERTNOATAES What in the of experiment, decinormal | sation q 7 2 No ee ae expressed aS} __ 380), g f a fraction | ane of normal sea-water. strength. 2 hrs. 16 h. 30 m. 1 Control | — | Division into two cells com- | Blastule. Twenty to twenty- mencing four cells in circumference 2) 05 ce 000025 |Sameascontrol .............. Blastule. Twenty cells in cir- | cumference. A few morule 3 0°75 ,, 000037 | Same as control ............... Same as No. 2 4 TL O) 5 0 0005 Same as control ............... Same as No. 2 5 oS y; 0:00075 | Same as control ............... Half morule and half blastule | 6 2:0 0-001 No division .........00.e.00008 Same as No. 5 Experiment 2, Brood No. 6. In this series blastulee formed with smaller number of cells in embryo. Amount of Amount of added alkali | Observations of embryos at stated times after commencement ry RUTTEN in the of experiment. decinormal Tati solution ee eerie expressed as 200 c.c. of a fraction sea-water. | 0%, noxmal 18 h. 30 m, 42 hrs. strength. Control — Four-celled, many morule | All stages to early blastule and a few early blastule i ce 0 -0005 Two-, three- and four-celled, | Morule, at all stages and morule 2 op 0-001 Many single - celled, and | Harly and late morula. Ex- stages up to early morule trusion of cell contents from cells By, 0 -0015 Nearly all single cells ......... No further development. Ex- trusion of cell contents 116 Prof. Moore, Dr. Roaf, and Mr. Whitley. (Oct... Table VIL.—Calcium Hydrate. Experiment 1, Brood No. 5. The lime-water added required for neutrali- sation of 50 cc., in one case 25:25 cc. of N/10 HCl, and in a second trial 25°28 c.c.; this lies so near 25 cc. that the solution was takem as N/20. | Amount of Amount of | jaded alkali | Observations of embryos at stated times after commencement calcium in the of experiment. hydrate in Tuti No. | N/20 solu- rae ag tion, added | ©*Prpssty as - Aaa ae) fraction sea-water. eee 18 hrs. 42 hrs. il Control — | Well-developed blastule...... No further development. Com- mencing degeneration 2 1 ce. 0:00025 | Blastule clearly more ad-| Late blastule, one or two vanced than control, and| moving showing more cell-division 3 2s 0 -0005 Same as No, 2 ..........2.0005 Same as No. 2, but more in motion 4 4 ,, 0-001 Well-developed blastule, but} Blastule, but motionless some deformed in shape 5 @ 5 0:0015 Morule, and occasional small | Blastule, but not so far blastulee developed as No. 4 6 8 ,, 0-002 Nearly all single-celled; but | Nearly all single-celled. Those occasional two-, three- or| showing development are five-celled group chiefly early irregular morule. Single cells show extended layer. 7 10 ,, 0. 0025 Single cells in many cases} No development; single cells | pear-shaped showing extended layer A second experiment gave similar results. inus. ial 7¢ etc., on Eggs of Ech 2 ves, Effects of Alkal 1905.] ‘squomtedxg pourqutoy ves ‘syueutttedxe 1oy}0 10g “T9¥VM-BAS Jo “0°O YZ UT plow oL1o;yoorpAy QT /N peppe Jo'0'0 F 4 Sutky “In990 09 ATortyUS paseed UOISTATP-[[90 TOIT 4¥ JIVAT] OY} “4[Nser TeTIUNIS B aARD ‘z -O N poorg ‘Z quowrodxyy ‘quoutttedxe 044 Jo pus oy} [gun pearosoad Tom poureuar Loy, ySnoyye ‘prow 043 Aq party ueeq Atjuoredde Sutaey stjeo 09 ‘eatyeBou sem y[Msex oy, “posoord 0% UOISTAIP MOT[e P[NOM plow eT4 JO WORSI[VANEU FI 9es 04 LapzO UI “peppe TeALe OL/N JO 9'9 ¢.g Sutavy roy, oy} pue ‘peroyfeun 4Ja] Suteq auo “9:0 QOT Jo suoKtod [enbe omy FUT poprarp sem 1 ON potsod siqa4y s][e0 PesT[I}1og O[SUIs [TV |""*"*** peatasqo yony |°* Saae poT[es-ay3urg |-""** x peT[ed-ep8utg |"**"""** parjeo-epsutg SZ00.- O “0.g L poeT[20-anoy pue -omy oumog o9e4s | ‘emaou pourtoy | «emBerty ~“peTfoo peT[20-44 Ste w[ntou ysed ouoyy | -j[ey pue apnaocy | -yyS1e pure -xIg | pue “exis ‘-amog |'""'''*': peyjeo-e[sutg 0200: 0 “ 0.F 9 PpeTl99 pedojeaep Moz SsoTUO0T}0UL B[N.1OU FTVTy -OM} MOT B “PoT[ED yng ‘Z ‘ON SB ourng | auIOg ‘enyselg | pur apnyserq Fyey | eproy_ | -opsurs qe ATceoyy | CTO0- 0 “oe G BTN.AOUE FT VY pe][20-m0F Z ON sv oureg |" “““paadasqo gon |"*"""**""poadesqo jon | pus #TNyse[q J[eA | pue peTjeo-seayy, 0100-0 “0-2 Ge GON SB oureg |" [Oa}TIOD sv OUR |*** oOT}UOD sv eUIBY |**" [OL]UOD sv oR |"** [oOTQMOO sv omEg C000. 0 4 Ob Ti e SSO] “UOTOW «“KNAsSeT_ |*"* [O1}WOO sv oTMBG |*** JoOIZUOD sB oUIRG |'** oI19ZUOD Sv OTMBY |"*' [oT]MOD sv oeg ¢Z000-0 00 G.0 z SSOTUOTJOT, *BTN1y {UO ULOAOUL sutaout SSO “svs Surouemmog | pidea ut ‘epnqseig | moy v ‘elnyselg | -Moyoy, ‘epnysetg |*** e348 peyjeo-mmoq —_ sconcecoy@) I “SIT OL, ‘WM OS “UT SP “SIT, 6% ‘Ud CT ‘YF ST ‘a OF “FP ‘YQSUeT4s [eUOU jo *10YBM-BOS JO a WOT4IVAT B SB 0° 0OZ 0} poppe vi | pesserdxe “uornyos ‘aorynyos OT/N ‘ON, ‘quounItodxe Fo qUSTIEDTETAULOD 1047" SOMIT] poqeys 4v soLtquIO JO SMOTZBATOSGC, peep i0 ARO: sophia soquiting iS e) =| fe) = ‘TL ON poorg ‘T quomttedxq ‘ploy oMopyoorpAH— TTA OTe, ‘SIOV HIIM SINSWIMTaXT 118 Prof. Moore, Dr. Roaf, and Mr. Whitley. [ Oct. 9, Table VIII.—Acetie Acid. Experiment 1, Brood No. 6. Amount of Amount of | “Jaded acid | Observations of embryos at stated times after commencement acetic acid in th of experiment. in N/10 in the P No solution, ee added to | °XPressed as 200 c.c. of a fraction sea-water. O: Dome 18 h. 30 m. 42 hrs. strength. 1 Control — Many morule and early| All stages to well-formed blastule. Some four-celled | blastulse groups 2 0°5 c.c. 0:00025 | Morule, occasional early | All stages to late morule. blastula. Not so far ad-|} Division irregular vanced as control 3 10) 55 0 -0005 Morule only. Irregular | Same as No. 2 division 4A 20 ,, 0 -0010 Early stages and irregular | Karly morule, and some late morule. Many cells not| morule. Cells breaking up divided 5 3:0 ,, 0 :0015 None beyond six- and eight- | Chiefly early morule, and yf § : celled stages earlier stages. Irregular division in earlier stages 6 4:0 ,, 0 002 All single-celled ............... No stages beyond eight-celled. Irregular division "/ 5:0 ,, 0:0025 | Allsingle-celled ............... All single-celled IIS) Effects of Alkalies, etc., on Eggs of Echinus. 1905.] “7[NSOL IV[IUIIS B OAVS PIO’ OTMOG.ATVO TATA yUoWLIodxe puooes W 4B sino00 YyMors Surddoys ut plow ormoqivo AIM yooyo oy4 484 0g orqno ut AjayeuTxordde puodsertoo pappe plow o1moqatwo Jo syunoweT o4y 4vy} yno poqutod eq “splov o1qa0v Lo o1LoTyooapAY YAIA sv [OAOT OUBS O44 OGL °0°0 GT pus “0° g.g “9'0 1.2 “0'0 Z “9'0 g.T “9'9 1.0 04 WOINTOS OT/N FO SeajouIyUE. BU 4I ‘sjuotItedxe coy}O oY YATA UoOstAedutoD 10,7 Be IGESTA OLB SOUBIQUIOUL UOTZeSITH -Iay pus ‘poasresoad [jam ove YIno1yy pepqquq S[[99 OUT, “WOISTATP-][90 ONT fe seee* WOISTATp-[[99 ONT [eee STTOO OTSUIS TTY $600: 0 sed plow ormoquep 4 duos poy[oa-x18 LIVING UV 4SoySt_, “WOISTATp G ‘ON 8B oureg | -]Ja0 4e sqydureqqe oM9 AO oUQ fe sypeo opsurs TTF G9I00- 0 “ oe 9 s][90 paqutouesag |-"""'**"''* WOIStATp-[[90 [euMaoM ON | " s]jao osuts [TV Aptva yy ZETO0- O “ 92 g woes afaout dn P92][99-0M9 [RMOIsB000 poyersoqUIstp [TV | -Woyorq Moy Y “Suryeadoqursiqy | ‘peyfoo - epsurs ye Ayrwayy 66000- 0 Oi P ouos sojusdey “qoazaed osory Jo Aut UOISIATP LeTMSoaay sosvys ATtva ut @fnxout | AToorvos pue ‘xmaou tomo | ‘poyfoo-.mofz ou ‘paqjan pedojeasp-ATpeq ‘dn-ueyorg |g ‘oN uvyy peouvape ssory | -e01q9 [eUOISBI00 “paT[eo-om J, 99000: 0 ET g KTNysvyq Surouour ouyWOS UBYyy | -UI0D OM9 IO OUG ‘JOLJUCO UT | UOISIAIP IeMBeaty “poT[eo poouvaApe ssoy = *B[NAOPT | ULY} apNacoM Fo Toqunu sepPeutg | -moz [BUoIsvo00 ‘paT[o0-omM J, €000- 0 ‘00 J, Zz xlNyseyq xpnyse[q pus apntoyy | A[rvo Moy v ‘xpntou Apoyo |:::98v4s poyjoo-amoz Te Ape NT -- Joryw0g T “SI SP “Sa GT Ul OF UF ‘poppe 194BA-Bos “Yj steqT4s [eULTOU JO prow o1uog.vo FO WONIVIT B SB OUINTOA SNUTU ‘194BAL pessardxe ‘aornqos -BaS JO 0°0 QO 0} | “xequinyT ‘quotttedxe Jo JUSTIO.MOMIUIOD 1047e SOUT poqeys 4v sokaqtuo Jo SUOT}RALOS nae as ea lds een es eee TONE oma betes Wee eae ae poppe joyunoury | Nf ¢600.0 Ut pow oTMOg.avo Fo yuNOUTy TEXTS OL/N 00 G.6 = '0°O NOT oAvs ‘10yvorput se urepeyyyd -jousyd pue FORN OT/N YIM UoTyNTos siya Jo Morea, ‘“poyeis soyMueNb oy UT TOJwA-vas ay 0} peppe pu peredord sem ‘ysnoryy “OO surtqqnq Aq pourr}qo ‘10jvM-vas UL plow oTUOqAvo Jo UONTOS W ‘PF ‘ON pooag ‘T quowtodxgT "‘plOV_oLuOog.TeO— NT TIBI, OO CN [Oct. 9, Prof. Moore, Dr. Roaf, and Mr. Whitley. 120 morjour ur Kueut pue Aurut ]o.14m090 uvyj poouvape 880'T == EATS UOISLAIp Ley “nsont yng ‘Zz “ON 8B ‘quoutttodxe JO gUSTUEDTOWLULOD 10478 STAT} pozugs 4x sokuquIa JO SUOTYZALOSGO XB NOSTOUT- oMUTRAS FO WOVE B SB possoadxe ‘mo1ynjos oY} UI [woruteyo poppe jo yunowy D onateCon) ‘e[Nyseyq pue apnazout | oueg -pazjeo - moz uvy} poacosoad WOTOUL UL Sso[ | oyep pure ATresp “LoAO puw | a0 -oerq} RuOIsLO HO®N O1/W TOO “aNyseyg_ | Moy wv ‘apyseig | -uomou ‘engserg | amoz yo sdnows opquyunog | -00 ‘poqfeo-omy AGoryo W S100: 0 0'0 @ ¢ s]]eo Joaquoo 9Z—PTS SdUOLOFUINIAT/) usyy ps0Uvape 880/T Sso_ | ‘[oryW0D ULyy poouvApe | ‘poT[od - moyz yeuoIS HOtN O1/Ww = “Guryedouesocy | -uoyou ‘apyseyg | exout AyawoyQ e@ngsetg | -vov0 pur -e0.ry9 omy, WW 100.0 00 4 s][99 8I—9T o0uereyumMo pe]jeo-.moyz WOOT UL SSOT SSoT | -aIQ “uorjou ur you ‘ay | Aqrx0fepy “poTjoo ‘an ysurg | “Worour ‘angseyg | -woour ‘eemaserg | -ngse~q pouttog-pom porto | -anoz pu ‘-oaayy ‘-om J, — Joayu0R T “Wu Og “| OS ‘SUL 8p ‘sa] Te “Wl Sh “Yl GT “Wl GF "YL ee ee ‘poureu Yysue.t4s IVMOITOUL-orAUARAS oq} JO -0°0 UL ‘poppe [eorurero jo yunowy “ON “MOIJeAJUGOUOD pu LOINYYsSUOD owes oY} Jo dn opeur spmy Aq poovjdor pue uoyeyuvoep Aq Aep yore posuvyo o10M spy oy} quourttodxe siyy UT “g ‘ON poorg ‘T quourtiodxa XSeiQe ‘gouvysqns yove Jo pesn Ssuteq AjUO suo14e1yUd0N00 I : i doy} IO OM} ‘SqTBVS plow pue oUI[eyye oy pur ‘plow ‘eye SoulpnpoUT epruL seM soda Jo poorq ous oYy UJI SOLIeS POUIqUIOD *v ‘splo’ pUB Sollee JO WOTJOV oY} YIIM OSTe uosTAvduI0D v UTeS 07 pUR ‘sI[Vs YUOIOTIP oyy UdEMJoq SR afqeIvduIoD sz[NSeI oY} OxVUI 0} JOpIO UT ‘QoURL SIYY Jnoqe ye wostrvdwmMod IOF poxooTEs OOM SUOT} -C1JMONUOD MOF V SplvVMIO}}e pu “BADGE popdOdeI Sploe’ PUR Soll[ey]e YIM syuotUIIedxe oY UT Sse “pozooTTe o.LOAd 8500 SuIdopOASp OY} YOIYM 4 poulUMeJOp SVM JVs YOvo Jo UOTyaIZUEDK0D Jo eSuvI oY} sjuowttedxe Areuturead uz ‘SLIVG UNIIVATY GNV GIOV HLIM SLINGWIYadXY 121 ul OSU 109} 9GT Liffects of Alkalies, ete., on Eggs of Echinus. 1905.] wotyoUt TOG UUM *[OayUoo uri poddosoxd “OR TNASBL eee “‘SulyBtoudseq, sree FUTqRtOUdO CT UOIJOUL OATJON UT Aunuc ‘apnysepg: SSO] “uotjout —‘ayN4st[ Eg S89] -uorjout = ‘aanystlg + poqynrouoda(y eens HOA TOTOTO(T SOLOS UL Toqtunu snotaord Av uvyy poedoyoaop Loyjog “apnysvyq podopoaop - TIO nieiayerareatiele poyerouesoqy voeeevene poyeacoueso( JouyWOS UT WRI pooura -pB OLOUL aTNYSVTG OULT, 6 PUR g ‘SON, ‘osuys vrntour possrd | sv poouvApR any os OALY TOUINU AOT[VUIS B | FONT ‘“ANOF ogoTdu09 qnq ‘apngsryq 0} sosvys [TV | “UL puR poTpoo-«mnoyy SOSTIS LOTLTVO UL WOTS poouva “‘TAIP aupnsorty cuoduys | -pt [[OAA “SWOISTATp “STU ONT TLOLTEM faapngseyq | anpndo.ct TOUS podoposop-][oM 09 sosnys [Ty | pur pay[9o - anoyy UOIS “TAIP avpnsonty — *[o.4y1090 IVF OS SOLOS UL UY} OOUOAOFUMDATIO | JO pooUVAPT 4SOTAT UI S[[00 OOP, ‘apNAseTG | ‘potted aoysny pur podopoaop- [Jo 09 sodvys [Ty | -amoZ “-ootyy ‘ow UOISTATP INGOT “UOAOS 09 dn sdnoad wepnsot UOISTATP AvpNsatty ‘osugs | -at puv paTpoo-amoz gncowt ApIva puosog ouopy | OST “poTfoo-omg AURAL Sodus TOIPVO FO s[[OO Ur 4soz -TURUL WOISTAIP ce~nsoatty *(S[[00 FZ ooUecozUMI.A10) THMGoaat Joajzuoo ur weyy podopoaop | ATpoyavut — worstATp LOYJMY YOnU oxy aNysvtq | -]JoQ ‘xoAo pur oy} Jo AuvpL ‘“Rpngsetq | payfoo-anogz wepnsoary podojpoaop-[[oa 09 sodvys TTY | ‘poTteo-o.mg [Le ALweo py UOISTATP 4U sydure9qv oatqcoqe Lup uodvys ‘wniou Apvwo puokoq | -stut Auvur ‘stfoo Aur AToorwog “apngsvyq ON | postpytoF opus TV podoyoaop S30 yo Tuo JF[BYy-ouo uayjQ ‘quourdofoaop jo | avpnsonty ‘popratp SODRYS ][V JV RMAC «vp | WeYy O[AUIS oLOTA “nSottt Ato A “apNASvTq ONT | peT[eo-0.y pur -op surg POYVUt-][OM WOISTATP IN 0900: 0 W 0800. 0 W S100: 0 W S100: 0 W 100-0 W ¢T00: 0 W 100-0 *OOH™N OT/W 0'0 OT “OOT™N OT/IN 00 9 ‘OOH®™N 01/1 00g HOM O1/W 00 g TOM OT/W 0'0 Z 1OH OT/IN 00g 10H O1/IN 00 Or [Oct. 9, 122 puofoq suorstarp Prof. Moore, Dr. Roaf, and Mr. Whitley. SuTyer IBNSOL ouMos pur 'OdH'tN — — -ouesep ‘@pnsoyy fo ** ATCO @BlAoW | peyjes - moz AGoryD W 400: 0 ‘0'0 O Suyerouedep ‘ey yeube ynoqez ‘OdH'*N — — -NYSBlq pus alo | apysetq ATrve 09 soseqs [TY | Pe[[eo-mMoZ puv -oAy, W 0S00: 0 0'0 OZ poultoF-]JOM pus repnSoa S][99 W909XI8 suyerouesep ‘ey | Aqrtofepy ‘soseqsquozeyip | 07 dn sdnoad ouog 'OdH"2N 02/1 — — “NYst]q pur BMAP | 98 VMysetq pure wxpNAOP | “LV[MSoaTt WOIstATp [EH W §z00- 0 0'0 OL, SUOISLATP IMSL pure oyoyduoouy *poyjoo-e0a9 puoseq suors “Tarp oyoqdutoo mez £19 A 'Od*H®N 01/10 _— — “seers Neaerouoseqy | “erqeyuno0o sdnoas [Ty |’ eT “ON se oureg W §200- 0 0'0 J 800848 I[® Ul WOTstATp «vpn sory uodeysstpy “Sq]99 'Ood*H®N OT/W — — “ees9es Hoqerouosay | @pnzout Aprve puodog ouoyy | pestyyteyZ o[suis [TV W &z00.- 0 00 G soseys surAout Sutaout SSOT Blnyselq | LoysIy Ur spjeo etour £o0*8N OT/IN MOF B ‘BlNASBIG | Mey wv ‘alMaseTg | -UoTouL ‘aTNyseTg | podofeaop-[jom 0, sesvyS TTY | JUG “TT “ON s¥B oureg W 0200: 0 00 g WOISIAIp Lepn Sac peyrey ‘(sqoo puoseq suorstatp 01} 0UL Sutaout | eouedeyumMoao) apnyseyq | cepnSort pur “peTpeo £9O0°eN OT/IL Suryeioueseq | ut you ‘apnyserg | [etoaos ‘a@pnyserq | podopoaop-[jom 09 soseys [LV | -AMoFz pue ‘90.19 “OAT, WW ST00- 0 0°0 @ 4 : , * e : : ; ‘oagty tod *TO4VM-COS a OF “TOS Sil, $7 sty TE uw cy 4 61 eet ULC) WoTyeayue0t0D Jo -0°0 Og 04 “quowattedxo JO FUSOULIDUSTIMM0D o9FB SOUT pozuys 4V soaqme JO SUOTPBALOSG OQ LVNOOTOUL-oMMUAV.AS JO WOToRAT B sv possordxe ‘wo1nyjos oY} UI [BoTUEYO poppe jo yunowy ‘peuvu Wsuec93s IV]NO9[OUL-ouLUTBAS oy} JO 0°O UL “poppe [eormero jo junowy 41 9T VI &1 ae IT “ON ‘panurjuoo—X 9qRI, 123 Effects of Alkalies, etc., on Eggs of Echinus. 1905.] Tensor ATU “apnctout poouwapy Jorqu09 URYY poouRApPB ssorT ‘apnaout ATTRe pur ‘peTjeo-tnoy pus ‘-oorty ‘omg cepnsetat L194. JO1jUOD TIRYY peouURApR oLOP “AVMSoItt outos fpoouvaps ATUEy BNA poouvape coyyiny ouou £ poT[eo ee.tt[y IvpNGe1at ‘payjeo-oay ooydutoo pue ojojd “ULOOUT [VUOISLIDQ ‘poTfed-epsurs [[e ATTeONT UOISIAIp AVNsemty ‘aparout AT.two pure ‘sdnors porfeo-tnoz pur ‘-eo1y4 ‘OMT, Jorgt090 Uv] poouUBApPY sseT “WOISTAIP Av[NSOIT KtoA ‘otout a0 sffoo 4ysta jo sdnoasd Ivpnsorit pues ‘poTjeo-eerty “peT[90-Oa\T, [O1}UOD WRI poouBApe OLOT, “B[N.AOT|] encour <.cedT | Wey, 994s poyjoo-moF ut ATPeyAted Mou pure ‘optssu0ye pouTUVEXe ‘TOTJUOD UBYY poouBape otout STsnNOIAgO ‘dnoad : poyjeo-WeeqxIs pe -4l[dIe [RUOTsB00O “payjeo-anoy oejeydmoo puv arjnset qe AytweNy ATxRO 09 peTjoo-oMg AeTNSettt pur ogoydutoo mo1f ‘soseqs TLV poouvape oL0]\, VY ‘poyjeo-ou avpnBex pue oqeyduroo qe Ayteo Ny -MMODUT OMY LO OUQC) pues «emery ‘poptap Apoyotdutoour ettog “pejes-omg AToryO JO s]joo WyIM sdnots poouvape ocour pur ‘peT[oo Moy pue ‘-oattg ‘OAT, pure -901T} OAOT UOISIAIp tensor Ata A *pa[[9d-MOF 10 -oa.tty} ojo[dutoour jRMoIsB00Q, °OOH®N OT/IN ‘SIq W 8100-0 00 § 8 S][99 POZIS-AVTNGOLAT TTA “eapN.Low AO) WEIR) WORD Ua HOM 01/1 ‘SUOISTATP IB[NSetIT ATOATsseOXT W ST00- 0 *0'0 g L PSTleotnoy PUB) -oecty, Key HOM OT/W W 100-0 ‘0°0 G 9 peT{eo-og ogoqd 10H 01/1 “poypoo-eysurs Te ATpeoroetg, W S100: 0 00 § ¢ sdnoid poT[eo-gy ste pur -«mojz oqe;dutoout ‘ezIs Ul IvNSotat pur aeMsed IOH OT/N W 100. 0 0'0 ¥. ezis puv odeys reso. HO®N 01/1 ‘HOISIAIP ep Seaat Aro A. W ST00- 0 *0'0¢ g WOISTATP LepUSertt 07 Louep ‘UOISTATP LOSI TeMotsvooo { popjeo-moyz HO®BN OT/N ‘JouUOD UY} poouBape oAOTA[ WW 100-0 00 & (6 *poyeo-omy TTR Aye NT — ouquo0y T ‘Ul OF TL ‘oagty cod "L04BM-VOS TOI R.14MEDTIOD FO *0°0 OOS OF TRMIETOW-oUMUAV.S owen yysu9.148 Jo WoTqoRAT v IvNOoPOUL-ouTUTVIS “ON ‘quoumttadxo JO JUOTLOOMOUMULOD 10}}B SOUT} PozRzs 4B SOATGUTO JO SUOTFLATOSGO se possatdxe ‘MoTynTOS olf} UI [BorMETD peppe Fo yunowy OT} JO Sacqgoutryw989 oIqno ul “pepper eoruleyo Fo JUNOTAY (OE 01 BT pUe LT O} T ‘SSY) MEATS suOVASNTTL OY} SUIYeUI OF posn o19A suonvavdeid ey pue poqiiosep sv ‘poureys pur ‘pouoroes ‘pexy orem quoutttedxo styy ur pedoreaep sokiqme oy, ‘6 ‘ON poorg ‘Z quountodxgy IDSC OENE [ Oct. 9, Moore, Dr. Roat, and Mr. Whitley. Prof 124 o aZIs UL IBMIOLIL BLOUL S[[90 PUL ‘UOISTATP UI pooueA “PB Sse] yByMormos yng ‘QT ‘ON 09 «ETI “poxyreud [Tom ‘OdH°®N 02/1 “xis “90143 “oq osje nq ‘poezfeo-anox Aqrx0fey WW SZ00- 0 *0'D 0 LI UOISTATP GT ‘ON UL wey poyteut atout sT[a0 Jo -T]29 UwlnSoxm rou ospy “eT ‘oN ur | quoweSucaie pus ezis ut Aqaepuseaty ‘sdnoas wey} peouBaps sse_ Yonut ‘pedojerep eavy | poyjeo-eeryy pus -omy ose {sT[e0 oxoUL Suture} OdH8N 02/1 Aoy} e1oYM BTNAoM pue ‘s[je0 e[suis oaopy | -uco Lueut qnq ‘sdnows pojjeo-amoz ur Ayoryy WW 0S00- 0 0°09 0G 91 1 pue g ‘SON 8B Os YonuT Os you 4nq 'OdH*8N 02/10 e[ntout peouvape Ape | “wwpnsexty *payfeo-xnoy ayoyduroour pure oyedu0g W §z00: 0 0°D OT cI Pete -9suIs [Te AT[vorovrd ynq ‘peTjeo-ee1q} pure 'Oa°HEN OT/W -OM} OFUT WOISTATP 4v sqdutoqqn wen saaat IecyN *jslal leis niala|o[ote sie(o/e elaine ies veleleieeivals es 1 N9aTtao-o[3uIs TlV W £200. 0 "00 J, FL e[n.OUL ON ‘poTpao-epsuts Aqraofvur qnq ‘parteo peT[eo-sext4 eqaTdutoour pus ‘Od°HEN OT/IN “O0LTT} PUL -OMF OFUT SUOISIATP Ae[NSeI L190 A | Ae[MFeLM [BUOIsvoIG “peT[eo e[surs [[e ATTeooRAg W $z00-0 00 G el [199 JO ezIs ww[Nsacat puv dnox8 Jo odvys IBMSIIT YAIA ‘epnacour ATTve osye * peT[e0-e01494 ®OO°8N OT/IN TT ‘ON UVy} poouvape ssory “a[NAOP | puB -omy [BLoAOs nq “poT[ed-tmogz aBpndex Luvp WW 0800: 0 0'0 g ZL OL ‘ON Usyy tepnset S][99 Fo Loquinu wesc] YIM pus ®O0°e8N OL/W aout puv ‘peouBaps otout AYYSITG ‘a#paop | ‘xv~nFoam Moz W “poTjeo-«noy xepnBoea [[v Ayxwo N TW ST00- 0 ‘0'0 g IL TeNSoeaIt S][99 JO ozI8 UI AqITeMSoAIt [BUOIStOIQ, “peed ®OOH®EN OT/W ApYysys yng “6 pue g ‘son sv ouvg ‘amaoyy | -9y510 pue -x1s Mey B f poT[eo-moz Te Aya N WW 0800. 0 00 OT OL *OOH®N O1/W 8 ‘ON UBYy peouvape ssoy AYSITg “apnTOpy |e Harao-anog aqe[duroo [pe Aqxwe NT W 0800: 0 00 9 6 “SIL g ‘Ul GF Ye ‘aaqtT od "1078 -BAS 10198139900 JO ‘0° QOZ 04 IB[N99[OUL-sUUTeIS ‘pemeu 44dueI148 jo moMourT v Ve[Nd9[OUL-euMTBIs ‘ON ‘quoulI4edxe Jo JUSMIEDMOUMULOO 10478 SoTL) poqeys qu soAIqUL JO SUOIYVAIOSGGO “panurjuogIi—TX Qe, 8B posserdxe ‘aorqnyos ey} Ul [BorTuTeT[O pepps Jo yunoury 9Y9 JO somjou1yUe0 o1qno ur “peppe [eormeys Jo yunowy 1905. | Fiffects of Alkalies, etc., on Eggs of Echinus. 125 Summary of Results on Rapidity of Growth. The following conclusions may be drawn from the observations on the developing eges in the living condition recorded in Tables I to XI, which are also illustrated by figs. 1 to 17, drawn from sections of the eggs developed in the experiment described in Table XI. The sections were drawn under a low magnification (61 diam.) with the Zeiss Zeichen-ocular. The eggs shown in the drawings had all developed, for the same period (8 hours) in the different media specified, from the same brood of eggs, and the eggs were subsequently fixed and stained as described above. 1, The extreme limits of variation of hydrogen and hydroxyl ion within which growth is possible are very narrow. Addition of 0:0015 M. of caustic alkali (see figs. 3 and 7) on the one hand or of 0:001 M. of acid (see fig. 4) practically stopping all cell-division. 2. A slight inerease in alkalinity favours growth and cell-division and at the same time tends to produce irregularity in size and shape of the resulting cells (contrast fig. 1 with figs. 2, 6, 8, 9, 10, 11, 12, 15, 16, 18). 3. In the case of small additions of acids no such favouring action is observable, but from the beginning cell-division and growth are inhibited (see Tables). 4. Increase in alkali above the optimum amount leads to increased and irregular nuclear division unaccompanied by complete cell-division. As a result the cells become multi-nucleated. The cells also become excessively irregular in size and shape (see figs. 3, 7, 8, 10, 15, 16, 17, 18). 5. On the other hand increased acidity leads in many cases to action upon the chromatin of the nuclei, so that in the sections the nuclei stain faintly and are comparatively few in number, and there is no proliferation of nuclei in the undivided cells, similar to that seen in the case of alkali (see figs. 4, 5, 13, 14). 6. The primary factors affecting the rate of growth appear to be the variations in concentration of hydroxyl and hydrogen ions. Thus all the caustic alkalies are of approximately equal power and there is little or no action of the Kation (see Tables and contrast figs. 2 and 3 with figs. 6 and 7). But in the case of the phosphates of the alkalies where the hydrogen and hydroxyl ion concentrations are comparatively low, there appears in addition to be a specific factor. See the marked action of 00025 M. NaH,PO, in practically stopping cell division, while 00050 M. of NazHPO, has a favouring action (contrast figs. 13 and 14 with figs. 15, Ho; 17). 126 Prof. Moore, Dr. Roaf, and Mr. Whitley. [Oct. 9, se) Q eS TON : an oy = we ASIN S Ona tae LP Sy, Q ae IN) Be & we ..0 oo” CB CQxpy as se ks S) eo VY F = SLD> e. @ D> @ > Fie. 11.—Growth of eggs in sea-water, Fic. 12.—Growth of eggs in sea-water, +0:0015 M. sodium carbonate. +0:003 M. sodium carbonate. 128 Prof. Moore, Dr. Roaf, and Mr. Whitley. [Oct. 9, Fig. 13.—Growth of eggs in sea-water, Fic. 14.—Growth of eggs in sea-water, +0:0025 M. mono-sodium phosphate. +0:0035 M. mono-sodium phosphate. @ Fic. 15.—Growth of eggs in sea-water, Fie. 16.—Growth of eggs in sea-water, +0'0025 M. di-sodium phosphate. +0:005 M. di-sodium phosphate. B pO Fic. 17.—Growth of eggs in sea-water, Fic. 18.—Drawings of cells in fresh condition +0:0075 M. di-sodium phosphate. developed in alkaline media, showing irregularities in size and shape of cells. 1905. | Effects of Alkalies, etc., on Eggs of Echinus. 129 Fie. 19.—Undivided egg in sea-water, Fic. 20.—Undivided egg in sea-water, +0°0025 M. di-sodium phosphate, +0:001 M. potassium hydrate, showing showing two dividing nuclei. several multipolar mitoses. Fig. 21.—Cell in blastula in sea-water, +0°001 M. potassium hydrate, showing multipolar mitosis. Fic. 22.—Cell in morula in sea-water, +0:001 M. potassium hydrate, showing asymmetrical mitosis—unequal number of chromosomes. 130 Prof. Moore, Dr. Roaf, and Mr. Whitley. [Oct. 9, Fie. 23.—Cell in blastula in normal sea-water. Fic. 24.—Cell in blastula, in sea-water, 4.0:0025 M. di-sodium phosphate, showing rod chromosomes, unreduced in number Fic. 25.—Cell in morula, in sea-water, +0:005 M. di-sodium phosphate, showing partial shortening of rod chromosomes. 1905. | Effects of Alkalies, etc., on Eggs of Echinus. 131 Fic. 26.—Cell in morula, in sea-water, +0°0075 M. di-sodium phosphate, showing shortening of rod chromosomes to dots, and reduction in number. Fic. 27.—Two cells in blastula, in sea-water, +0°005 M. di-sodium phosphate, showing partial shortening and reduction of chromosomes and protrusion of spindles, as if polar bodies were being formed. Fig. 28,—Egg in two-celled stage in sea-water, +0°0015 M. potassium hydrate, showing shortening of dividing chromosomes to dots and reduction in number. 132 Prof. Moore, Dr. Roaf, and Mr. Whitley. [Oct. 9, Fic. 29.—Cell in blastula, in sea-water, +0°0025 M. di-sodium phosphate. (See text, p. 134.) Pathological Nuclear Divisions. The irregularity of growth in sea-water to which alkalies or alkaline salts have been added is accompanied by characteristic irregularities in the mitotic divisions, which are illustrated in figs. 19 to 30.* Multiple Nuclec—tThe divisions in the cytoplasm occur less frequently than in the nuclei, so that the cells become multi-nucleated (see figs. 2, 6, 8, 9, 10, 15, 19, 20). This occurs both in undivided and divided eggs. In many cases two or more such active dividing nuclei are found in the same cell (fig. 19). Multipolar Mitosis—This occurs both in single-celled eggs and in the later stages (see figs. 20 and 21). The chromosomes are unequally distributed to the different spindles. Asymmetrical Mitosis——This frequently occurs in the various alkaline solutions, and gives rise to unequal nuclei. The cause of the unequal number of chromosomes appears to be that some of the chromosomes are carried to * The figs. 19 to 30 were drawn with the Zeiss Zeichen-ocular, the finer details being filled in as realistically as possible, under a Leitz =; oil immersion objective. The magnification measured by the stage micrometer was 790 diameters. 1905. | Liffects of Alkalies, etc., on Eggs of Echinus. 133 one pole without previous division. (See fig. 22, and lower spindle in fig. 19.) Changes in Chromosomes and Reduction in Number.—The short, rod-shaped chromosomes of the organism which we have been examining (Hchinus esculentus) do not show typically in its normal maturation or maiotic divisions the appearances described by Farmer and Moore* and others in maiotic division, and by Farmer, Moore, and Walkert in cancer. According to Bryce,t the main changes in Zchinus esculentus during maturation consist in a marked shortening and thickening of the chromosomes with reduction in number. Each chromosome divides into two short curved rods with spherical enlarged ends, giving rise to an appearance similar to tetrads. But at no time are there any rings or other irregular figures, as described in so many other organisms, and there is no true tetrad formation. R. Hertwig§ has observed in the case of Echinus a formation of bodies described as resembling tetrads, as a result of addition of dilute poisons. In our spe¢imens of eggs grown in dilute alkali, we have observed appearances in many cases similar to those shown in the drawings illustrating Bryce’s paper. As the amount of added alkali is increased, there occurs both a shortening and thickening of the chromatin rods and a reduction in their number. In some cases the appearance of rods is lost entirely, and the chromatin becomes arranged in minute masses resembling tetrads. These changes are illustrated in figs. 24, 25, 26, 27, and 28. Thus in the three strengths of di-sodium phosphate solution a shortening of the chromosomes in fig. 25 is seen as compared with fig. 24, and in the strongest solution, fig. 26, the rods are reduced to dots and the number is decreased to approximately one-half. In fig. 28 some of the chromosomes present a tetrad-like appearance. These changes, while frequent in occurrence, are not seen in all the dividing cells in any strength of solution, but occur in increased number in the stronger solution. A peculiar appearance is represented in fig. 27, in which a decided protrusion of one pole of the spindle in two adjacent cells of a blastula was observed beyond the cell margin, as if a polar body were being formed. Although it is difficult to count the chromosomes accurately, the reduction in number is obvious on contrasting the weaker and stronger solutions. In a fair number of cells, especially in the di-sodium phosphate solutions, a peculiar arrangement of the chromatin is observed, which is illustrated in * Q. J. M.S.) vol. 48, 1905, p. 489. + ‘Roy. Soc. Proc.,’ vol. 72, 1903, p. 499. t ‘Q. J. M.S8., vol. 46, 1903, p. 177. § ‘Sitz. Ber. Ges. Morph. und Phy.,’ Miinchen, 1895. Quoted from Wilson, “The Cell in Development and Inheritance,” 1904, p. 256. VOL, LXXVII.—B. 134 Prof. Moore, Dr. Roaf, and Mr. Whitley. [Oct. 9, fig. 29. The entire chromatin of the cell is seen to be arranged in two groups of circles at a distance apart corresponding to the usual distance of the centrosomes of a spindle, and in some cases, as in fig. 29, the remains of achromatic fibres are indistinctly seen between the two groups of circles. The cytoplasm surrounding each group of circles is distinctly lighter and freer from granules than the rest of the cytoplasm. In structure each circle consists of a thin ring of chromatin showing distinct thickenings, about four in number, arranged approximately about equidistant around the circum- ference. The interior of the circle is clear, and both the outer and inner | border are clearly marked. In some cases this arrangement of the chromatin in circles is also seen where there is only one group of circles as in fig. 30. Discussion of Results and Summary. Our attention was attracted to the study of the effects of small variations in reaction upon the growth of cells from the bio-chemical point of view, as a result of the observation that in malignant disease no hydrochloric acid is in general secreted by the gastric glands, no matter where the malignant growth is situated, which pointed to an increased alkalinity of the plasma. In the course of our investigations upon the rate of growth of the cell, when microscopic examination was made of the cells in the fresh condition, we were struck by the marked irregularities in size and shape of the developing cells in the alkaline media illustrated in fig. 18, which is drawn from cells in the fresh solution developing in sea-water, to which di-sodium phosphate has been added, and also by the marked tendencies to nuclear proliferation. This led us secondarily to a cytological investigaticn of the cells when fixed and stained to show nuclear division, as a result of which we have found the irregular forms of mitosis described in the text. These atypical divisions, which have been produced by variations in the medium similar to those which occur in the blood in cases of malignant disease, closely resemble the pathological divisions seen in the growths of malignant disease. The results of our experiments and their relationship to the processes in malignant growths may be summarised as follows :— 1. In nearly all cases of malignant disease the secretion of hydrochloric acid by the gastric glands is stopped or greatly reduced, and this effect is not due to local conditions in the stomach, since it occurs wherever the growth is situated ; but is due to a change in the distribution of salts in the plasma whereby the alkalinity is increased or the concentration in hydrogen ions diminished. 2. Addition of small amounts of alkalies or alkaline salts, such as di-sodium phosphate, to the medium in which cells are growing and dividing, 1905. | Effects of Alkalies, etc., on Eggs of Echinus. 135 causes at first an increase in rate of growthand division, but as the amount is increased, there appears a marked tendency to irregularity in size and shape of the resulting cells. Nuclear division becomes in advance of cytoplasmic division, so that the cells become multi-nucleated. As the alkali is further increased, both cell division and nuclear division are stopped. 3. Accompaning the increased stimulus to nuclear division given by the dilute alkali, there are seen many of the atypical forms of mitosis described in malignant growths. The variations from the normal illustrated in the drawings are: (1) multiple nuclei in the same cell in active division; (2) multipolar mitosis, occurring both in the single cell stage, and later in the development of the organism; (3) asymmetrical mitosis, leading to unequal distribution of chromosomes to the two daughter cells; (4) reduc- tion in length of the chromosomes as the strength of alkali is increased until the chromosomes appear as rounded dots, and accompanying the reduction in length there is also a reduction in number to about one-half the normal; (5) in certain cases the chromatin becomes arranged in circles, each of which shows a number of thickenings. The circles are arranged in groups in the cell, and appear to represent a stage in the anaphase, the groups being placed at about the usual distance apart of the centrosomes, and traces of the achromatic fibres being occasionally visible. 4. No such increased growth or stimulus to nuclear division is given by varying the normal reaction of the medium in the opposite direction, by the addition of equal small amounts of acid. From the beginning the minutest amount of added acid has an inhibitory effect upon growth and nuclear division. There is no nuclear proliferation as the amcunt of acid is increased, and at a very slight amount of increased acidity all attempt at cell-division ceases. In the fixed and nuclear stained preparations cell-division figures are absent, and in the resting nuclei the staining power of the chromatin is decreased, so that the nuclei present a washed-out appearance contrasted with the normal nuclei or those of organisms grown in dilute alkaline solution. 5. The extreme limits at which life and cell-division are possible lie close together, indicating that the cell is very sensitive to even slight changes in the concentration of hydrogen and hydroxyl ion concentration. Thus the addition of so little as 0:0015 M. of either alkali or acid to sea-water practically stops all growth. On account of the presence of phosphates and carbonates in the sea-water the change in hydrogen and hydroxyl ions caused by such additions cannot be large. [Note added October 24.—Since the paper was written we have had an opportunity through the kindness of Messrs. J. E. 8. Moore and C. E Walker L 2 136 Lifects of Alkales, etc., on Eggs of Echinus. of the Liverpool Cancer Research, of examining the appearances represented in figs. 29 and 30 under a specially high magnification of about 3000 diameters. The magnification was obtained with a 10-inch tube, 3 mm. Zeiss apochromatic objective, 27 compensating ocular, and with monochromatic green illumination. Messrs. Moore and Walker have pointed out to us that these bodies, which appear as described in the text when seen with an ordinary ;),-inch oil immersion, are, when seen with the 3000 diameter magnification, really spheroidal bodies consisting of an inner mass or vesicle which scarcely stains at all, over which ramify filaments of deeply staining chromatin. Further examination with this magnification also demonstrates that the chromosomes represented as dots in fig. 28 are, in reality, also vesicular chromosomes. Even with the ;4-inch oil immersion we had been able to see a well-marked black line around the periphery of each dot, but had been unable to determine whether or not this was merely an optical effect. The 3000 diameter magnification, however, clearly shows that the structure is the same as that of the larger masses in figs. 29 and 30. It is interesting that these minute chromosome masses occur upon the spindle in exactly similar fashion to the normal rod-shaped chromosomes, and that wherever they occur there is a diminution in the number of chromo- somes. In many cases an occurrence of both rod chromosomes and spheroidal chromosomes is observable upon the same spindle as if conversion had only partly taken effect. Where conversion of the entire number of chromosomes into the spheroidal variety is seen upon the spindle, the number of such chromosomes is usually approximately half the normal number, as in fig. 28. But when they are found after separation has taken place, as in figs. 29 and 30, the number is very variable, and as the number diminishes there is a corresponding increase in size in the individual chromatic vesicles. We are informed by Messrs. Moore and Walker that bodies presenting a similar appearance are seen in malignant growths, and also under other pathological conditions, and in certain normal tissues. The question of the relationship, if any, of the diminution in number when such spheroidal chromosomes are formed, to the reductions occurring in normal maiotic division, and in cancer cells, we prefer to leave at the present entirely open, since we have no evidence that the diminution in number is effected by the same means; and we merely point out the diminution of the number of chromosomes upon the ‘spindle as a result of increasing the alkalinity of the medium, as an interesting and suggestive fact.] 137 A Note on the Effect of Acid, Alkali, and certain Indicators im Arresting or otherwise Influencing the Development of the Eggs of Pleuronectes platessa and Echinus esculentus. By Epwarp Wuit ey, M.A. (Oxon.). (Communicated by Professor W. A. Herdman, F.R.S. Received November 14, —Read November 23, 1905.) I. Errect of AcID AND ALKALI ON THE EGGS OF PLEURONECTES. While working last spring at the Port Erin Biological Station on the effect of acids and alkalies upon the development of Echinus eggs, it was suggested to me that it might be interesting to try the general effect of similar solutions upon some other type of organism. For this purpose the eggs of the Plaice (Plewronectes platessa) were selected, as they were to be obtained in abundance from the fish-hatchery attached to the station, but, as time pressed, it was only found possible to experiment with one acid and one alkali, and decinormal solutions of hydrochloric acid and sodium hydrate were accordingly used. In a pond attached to the hatchery numbers of plaice are kept in the Spawning season, and the surface of the pond is skimmed each morning for the purpose of collecting the eggs, which are then placed in the batching apparatus. Three batches of these eggs of different ages were taken for experimentation—those freshly skimmed from the pond, those which had been removed two days before and had remained since in the hatching apparatus, and those taken 10 days before and similarly treated. Some few eggs always escape the net in the process of skimming, and are taken in the catch of later days, so that, when 10 days old eggs, for instance, are spoken of, what is meant is that none can be younger than that, although some few may be slightly vlder. Such as showed obvious differences in age from the majority of each batch were removed. The eggs were treated in a similar way to that employed in a research carried out at the same time on Echinus eggs,* namely, they were placed in small batches in a number of tumblers, each containing 200 c.c. of sea- water, to which measured amounts of the decinormal solutions of acid or alkali were added. A summary of previous work on the effects of acid and alkaline solutions upon development will be found in the paper above referred to. * See preceding paper. 138 Mr. E. Whitley. Effect of Acid, etc., on [Nov. 14, The accelerating effect of small quantities of alkali on growth observed by Loeb in the case of Tubularia,* and the eggs of Arbacia,t and in those of Echinus, in the research mentioned above, was not noticed, but attention may be drawn to the very powerful results of even a small disturbance of the chemical equilibrium, which can be seen from Table I, where it will be observed that, after six days, 4 c.c. of decinormal acid or alkali in 200 cc. of sea-water (v.¢., a five-hundredth normal solution) produces a death-rate among fresh eggs of 75 and 44 per cent. respectively, against only 5 per cent. in the Control. In the experiments above alluded to with the eggs of Hehinus esculentus, it was found that acids and acid salts above a very small concentration pro- duced more deadly effects than corresponding quantities of alkalies and alkaline salts, and this was also found to be the case with Pleuronectes, and can be well seen in the same table (1) for strengths of 2°5 c.c. and upwards, especially with the younger eggs. Probably three factors enter into the explanation :— (1) The fact that part of the alkali added is immediately thrown out of solution as insoluble hydrates or carbonates. (2) Alkali is constantly being used up to neutralise the acid products of metabolism—chiefly COs. (3) According to Loeb, the presence of weak alkali assists the absorption of oxygen by the organism. If this be the case, the eggs in the acid solutions not being able to absorb oxygen so readily as the others, might probably be less resistant to the action of the reagent. The tables seem also to show conclusively that the younger eggs are far more sensitive to the action of the acid or alkali, and generally to the influences of their environment than are the older eggs, or newly-hatched larvee, which are, indeed, extraordinarily resistant. Thus, on referring again to Table I, it will be seen that a very large percentage of the older eggs survived 7, and even 11, days’ treatment, whereas, of the fresh eggs, over 25 per cent. in the Control, and a much larger number in all the other cases were dead within 10 days. In this table the effect only of very small quantities of acid or alkali is given (five-hundredth normal and under), but experimentation with somewhat larger amounts gave the same result as shown in Table II. Thus, it needed only 6 c.c. of decinormal sodium hydrate to kill all the fresh eggs in four days, but more than 8 c.c. to kill all those of two days old in the same time, while, by the end of that period, 30 c.c. of alkali * ‘Univ. of California Publications, Physiol.,’ vol. 1, 1904, p. 137, and ‘ Arch. f. gesant. Physiol.,’ vol. 101, 1904, p. 340. + ‘Arch, f. Entwickelungsmechanik,’ vol. 7, p. 631. 1905.| the Development of Eggs of Plaice and Echinus. 139 which was the greatest strength employed, had killed only 65 per cent. of the 10-days-old eggs. Acids, above a small concentration, had a remarkably stronger effect—in one day even 10-days-old eggs being killed by 10 c.c., and the two other batches by 6 e.c. When working with the larger quantities of alkalies, it was very difficult to tell exactly when an egg was dead. In the first experiments the point was taken at which the egg began to become opaque, but, as soon as the percentage of alkali present exceeds a very small amount, precipitation of calcium and magnesium hydrates takes place, which renders observation of such a change in the transparency of the egg difficult. Loeb, in his experiments on Fundulus,* finds that the precipitate itself acts injuriously upon the eggs, and that, if this be filtered off, the eggs will live and develop in much stronger solutions than they would otherwise do. The process of filtering off the precipitate before placing the eggs in the solutions was not tried in the course of these experiments, they having been carried out previously to the reading of Loeb’s paper. Hitherto, it has always been considered that the young larva, on first hatching from the egg, enters upon the most critical stages in its career, and is at that time most susceptible to external influences, but the experi- ments here carried out appear to show a resistance steadily increasing with age, and that, even after the rupture of the egg-capsule, the young larva is, at all events, no more susceptible than just before that event. It may be suggested that this steadily increasing resistance is due to the gradual development of the epidermal cells, which would form a protection to the young embryo more or less impervious to the surrounding solution. Tables III, IV, and V give the actual experimental data on which Table I is founded, and of which it is a résumé. Table VI is introduced with the object of laying emphasis on the statement as to the resisting powers of the older eggs. It shows the percentage of these eggs which succeeded in hatching in spite of the very unnatural conditions (stagnant water, possible overcrowding, etc.) in which they were placed. Incidentally may be noticed the very deadly effect of “di-methyl,’ to which attention is now to be drawn. I]. Errect oF INDICATORS ON PLEURONECTES AND ECHINUS. The effects which di-methyl-amido-azo-benzol and phenol-phthalein produce upon living organisms, as illustrated by the eggs of Pleuronectes and Echinus esculentus, were accidentally discovered in the course of these experiments. These indicators were originally added to the contents of some of the * © Arch. f. Entwickelungsmechanik,’ vol. 7, p. 631. 140 Mr. E. Whitley. Effect of Acid, ete., on [Nov. 14, tumblers, to show any changes in reaction that might take place during growth. When it was observed that they had a specific action upon the eggs, a series of experiments was undertaken with them, the results of which are shown in Tables VII—X. In all the experiments, except those recorded in Table X, two drops of the indicator were added to 200 cc. of sea-water, or sea-water plus varying quantities of alkali, in a tumbler. Table X records the results of varying the amount of phenol-phthalein employed. It will be seen that, although the indicators were made up in alcoholic solution, the amount of alcohol added in each case to 200 cc. of liquid, was quite insufficient to materially influence the result. The experiments without indicators, recorded in Tables VIII and IX, are included for purposes of comparison. It will be observed from the figures obtained that dimethyl is very deadly to the eggs of Pleuronectes and phenol-phthalein innocuous, while the opposite holds good with Echinus, the dimethyl having, if anything, a favouring effect on growth, and the phenol-phthalein being very injurious. Before killing, phenol-phthalein appears to be very effective in producing irregular divisions. The dimethyl was readily absorbed as such by both the organisms, staining them a deep yellow, so there can be no question as to its having thoroughly penetrated the tissues. So far as can be ascertained, this specific action of indicators has not been noted before, and no explanation can be given of the fact of the different indicators affecting the two organisms in exactly opposite ways. Tadpoles in tap-water, to which the same, and even much larger amounts of these indicators had been added, appeared to be totally unaffected by either. It might be of interest to repeat the experiments with other organisms, and with other organic compounds not known already to act as poisons. Summary. (1) The amount of variation from the normal concentration of hydrogen and hydroxyl ions in sea-water which the eggs of Pleuronectes will tolerate is very small. (2) A disturbance of the equilibrium towards the acid side is much more fatal than the opposite. (3) A progressive development of resistance to an unfavourable action of the environment takes place in proportion to the age of the eggs. (4) Phenol-phthalein is deadly to the eggs of Hchinus esculentus, but harmless to those of Pleuronectes, while dimethyl quickly kills the latter, and appears, if anything, to have a favourable influence upon the develop- ment of the former. 1905.] the Development of Eggs of Plaice and Echinus. 141 My best thanks, in conclusion, are due to Professor Herdman, Ren: to whose kindness I am indebted for the material for these experiments, and for permission to work at the station, and to Professor Moore, for his kind and valuable criticism and assistance throughout. Table I—Comparison of Percentages of Deaths in Fresh Eggs. Eggs of 2 days old and eggs of 10 days old, with varying quantities of deci- normal NaOH and HCl. After After = After | After After After 6 days. | 7days. | 7 days. | 10 days. | 11 days. | 11 days. Fresh 2days | 10days | Fresh | 2 days 10 days eggs. old. old) ‘eggs, |, old: old. | | Control, 200 ¢.c. sea- 5:0 0 2°0 25:0 0 2°4 water. (9 days) 1-0 c.c. decinormal NaOH 4°3 0) 4-2 | 1°5 » "2 9°6 = 0 29 | 2°0 » 5 25 0 ) 0-9 | 2°5 p es 18-2 2°6 3°0 31°8 2°6 3°0 20 BS 27°3 12°7 2°5 31°8 14°5 2°5 4:0 op a 44°0 250 9°4 48 -0 48 2 26 °9 (10 days) 1-0 c.c. decinormal HCl...) 10-0 3°3 2°2 15 7 6 ners 0 8-0 4°8 20 a aa aa HSB eG 3-0 15 2°5 op BS Seq. LODO) 20°0 3°3 28 °6 26-2 3:0 op op -..| 33°3 39 °3 Thchy/ 33 °3 42 *4 4°6 4°0 » » 750 39 *4 18 750 39-4 5-5 Average, exclusive of | 22°9 | 12°8 32 41-4 28-9 9:9 Control | | 142 - Mr. E. Whitley. Effect of Acid, etc.,on — [Nov. 14, Table I1.—Actual Number of Deaths in Fresh Eggs. Eggs of 2 days old and eggs of 10 days old with larger quantities of acid and alkali than were employed in the experiments, the results of which are shown in Table I. 40 eggs in each tumbler. | | Fresh eggs. 2 days old. 10 days old. | | Ist | 2nd | 3rd | 4th | Ist | 2nd} 8rd | 4th | Ist | 2nd | 8rd | 4th day. | day. | day. | day.| day. day.| day. | day. | day. | day. | day. | day. Controleennnercscrnt 0) iL 2 3 1 il 1 1 0 0 (0) 0 3c.c.N/10NaOH.... 0 | 2 | 11 | 25 | 4 5 55 seal) 24 || aly | HY) 5 49 5 | 0 5 | 34 | 36 0 1 1 iL 0 0 0 2 6 ; | 2) 1-29) | 36) all | Siac’, TO Pas PO 2 |p | ae 10 $5 - IS HS eS eS eS SK ] KH LS 0) 1 2 | less | | | | than | 10 15 x3 55 | — — j;— J|— | — J — — |'— 0) 2 3 | less than | 10 PO ashes alee — {|= |] =] j— = 04) eee 30 oy) ” rae Fal ae aa aa ray aaa en ® i B 26 | — — 3 ¢.c. N/10 HCl...... @ | 1 2 7 Wy deg RR RP ana oD oi Of wy | Be nee eee [75/8 |i som tee | 9 | 20 | 23 Strengths of 6, 8, 10, 15, 20 and 30 c.c. acid were also tried, all being fatal in less than 24 hours. The figures for the 2nd, 3rd and 4th days are cumulative. 1905.| the Development of Eggs of Plaice and Echinus. 143 Table ILJ.—Actual Number of Deaths among Fresh Eggs. | No. | 2nd]| 3rd | 4th | 5th | 6th | 7th | 8th | 9th | 10th of eges. | day. | day.| day. | day. | day. | day. | day. | day. | day. | @omntrolieecers 20 1 1 1 1 il 1 2 3 5 | | | | 1:0 cc. N/10 NaOH | 23 0 0 0 1 | Le ae i 21 @ Weal 1 2 2 0! 5, - 20 1 4 4 5 5 AG, i 22 4 | 4 4 4 4 6 6 6 | ia acon °.,, i Tp We aky NPGS 5 5 6 6 6 7 it ASO 2 4 25 Shs lp atl i | ob 12 12uo| ok | 1:0 ce. N/10 HCl...| 20 1 yD Dew 2 3 2 | Leas f 21 Oi @ Oo ) | BO! 4, ys 23 By | Dah e | | BSitey |; < 21 2 3 3 3 4 5 5 Sie 350. * 24 eeOic | 6 fi 8 8 8 Shales AO, ,; mn | 8 20 Seu iguow lp 15: 15 15 15 15 vey | || ais | | | The figures in each column in this and the following two tables give the total number dead by the day in question. Table [V.—Actual Number of Deaths among Eggs 2 Days Old. | | | No. of | 2nd} 3rd | 4th | 5th | 6th 7th 8th | 9th | 10th | 11th eggs. | day. | day. | day. | day. | day. | day. | day.| day.| day. | day. | | | | } j Gontroline: c.sccne ies 36 QO WON Oro oo 0 ) ft) 10cc.N/10 NaOH | 28 OW OF Cio Ole 15 i % 23 OP ON Ol O 1 OT oO PEO, 5 a BIO 1 O | OW OA OW Oo 25 if eS 38 TU ey Pega dal ogre al alec. Wy Kalil |i a 1 1 50 awe i. 55 OU OO) NO alae seh 07 ele aS 4:0 5 56 euch ere ee Ome On lee 235) (623) 523,05 27 i — | | j | | 10c¢c.N/10HCl ....| 60 | 1 Phe) a ae 2 2 | Tbe be rl, Db OT lhe Pounce NHR) Salta geal | 2 Ae vaca WBS Looe Teal ae Wee! Nie Bey, el 280 SEOs ee OnatOM eo imtGwe | tz | 177 | 19. | 21 Oe eon On 10) LOM LON 13) eis i4a 4) 14s | 14 Ue aetna 95 | 28 | 23 | 23 | 26 | 81 | 84 | 84 | 34 | 84 | 34 144 Mr. E. Whitley. Effect of Acid, etc., on [Nov. 14, Table V.—Actual Number of Deaths among Eges 10 Days Old. No. of | 2nd | 3rd } 4th | 5th | 6th | 7th | 8th | 9th | 10th | 11th eggs. | day. | day. | day. | day. | day. | day. | day. | day. | day. | day. Control ......... 245 0 0) 1 4 4, 5 6 6 1:0 c¢.c. N/10 NaOH | 165 0) 2 4 6 | 6 7 PD eee i, 105 ©}! P| ares pee sti es 20. a 110 1 Tew Peale tee rl 1 Rae i 100 le Ome tal st) i oe BDO) g ‘4 80 CE ORAROM it ha Le Sales 2 2 AiO! - 160 Ue WS | TS SO Peet ule 10 c.c. N/10 HCl 135 Nap els | a les | ree ‘ SOW Oke |S | 2 |e | 2°0 x) 0 155 0) 0) 0) (0) 0 iL Bi ers cool 120 0 1 2 Silints: A | 3:0 x 0) 175 0 3 3 3) | 3 3 3 5) 6 8 Ai Oun * 110 0: ), ila |) cepa SR Ma ee Salers al 85, 5 6 Control) .scs asses eeen 98-4 +1:0 ec. N/10 NaOH ...... 96-4 15) fe Wie 99-0 2:0 Ae 2 oer 99-1 2°5 bo nt eae 99-0 30 : AA es BOCaT 4-0 Pe RNY Scitinca.5 95°6 stale OR CsCoe NI Al OMEN CR ee neces 98°5 1s SENG. avis okeeoere 96:2 & 2:0 ett TR esad ae 100-0 2S) Sub Rt gba eset dene 97-5 30 a yA ide Me Ay) 98°3 4:0 TT Se can ae 98-2 Dime thyly Ae jae en atone 39°7 Phenol-phthalein ............ 98°8 145 the Development of Eggs of Plaice and Echinus. 1905. | ie i a8 + ‘QSBqIS [BAIV] OY} poover ssdo osoyy JOG » ! | | seal 6-2 16.2 |%-% |0-% \0-2-|8-0 |r-0 | 0 ;—|—|4 |4 ]9 |s |9 |2 | Nine Nee Nae Wea lor iene | Of O— 1B Ie |e yi it |r fe | O0T/%-96| 4\8-09|2-68/9-6 |%-8 |%€ |Z | — | £9 | 09 | 9F | 88 | 93) 9 | & | © | o|-01 ©] oO} ©} O| ©} O10 |O }o Jo jo jo Jo |o Oe ane Ph el Bl Ge | Bd Bi 6 IB WB IB eae eae ee a QOL F-16 |9-82|F-12|F- 11/29 |83 | 0 | 0] 0 | OL) %9) 03) 9T;8 |% |% JO oe | oil ol @ | @ | ©) @ | oi) 6 lle le fe Wir pe ie pa ee Ie! Of O1 ©} O} ©} O]- Oj) |L1O 10 10 |o }o 06 | 09,09! SF OF sg! GZ QT] ST|—|8I|3E|/2t|6 js | 4/8 | He | H H | | oO co @ ~v lor) or co ies) bo e (>) ve) lee} wT lon) Ou e Pie |EIS (Fl EISI SIE EE ele fF El) Ee pees eal [eteucere eae ise Ge le eee eee od | ee esa sel a ° . . oS . 5 . + 5 e . . . oy | | —Jo pula 4v prop jo osuqueor0g —Jo pue 4v prop 1equinyy ono aon oon Non ONO NOS ‘skup Z Jorym0D ulepeyyyd-fousyg+ jou Jory a0K uleyeyyqd-[ouey gy Lqgourg yoryu0D urereyyyd-fouceyg veneee eee if Aqgoutcy “L0JVOIPUT pro shep OT pro sep g soso YsoLy "yoyeg. ‘so}oouOIMe[g Jo sssq oy} uodn s10zeorpuy jo yoo oy} SuIMoYS—TTA %q%L, ~ on suryetoues > “op yok 40N io) sut A, -ye.toueda(T OATq0R Ato A Effect of Acid, etc., on 10 PIL Mr. E. Whitley. prop ‘poouvapre [To AA “ ce prep gnq ‘surpoeosead UBYy} poeouUBApT OTOP peed if3 it3 sur -yRiouesop pur prog Surpooead wey poouvape oxou pus PoULLOF c94430q yng ‘prog surpeooard se peoueape os jou ‘peep ¢ TL 092] ‘oanoy (13 (19 Sut “yetoussep pus prog “Ut ST “Y 16 SUTULLOF. “ it9 BTNASV) if3 “ ce ce aATqQOU SUTAOUT ATA STS “4p Ae Moy V (¢) poyusdeo OATIOB -u9 OULOS surtout Aparwy “49 Aqaueq| “« “ jeaT7Z08 OULOg] + TRataAeg eaTqoe A10A jou ia") Ayaan (vs e cc ce “ce its repnUeIs “SurUL surAouL ‘peep “1 Ajaegy| -wums “gq |“ “ joutos “g'|: ~» seenee 4 6 oulos pure ‘g “ cc “ce (Ts S89T bee eee nee it3 -uoout “¢ . soreeeees QATIOG “6 «cc 6c 117 its 19 6s . reyn -UBIS ‘pvoep T[v “19 SuIoueUULOD s § Peer Ay,sout “6 Apregy its “cc “e 73 peep Apysour “49 | guouteAout surAoUL sutouemutos “g} pidea “g | oaton [Ty | Moy @ “og WHOS TOL | WSh TIP) AST 16Z/ MOS ‘THz “104 M-BOG cole T “* eamgaserg $1 «x OCP “YLT aout pur ‘g ‘9 ‘fF OL pus 8 oos “Sot PF p oaoge S* pus “pF 8 pus g [BuoIseo00 pure “p 19 cc * pe][92 "8 rt ‘poToo-p + “" Bp MOTB pur ‘7 see eeeeeeeee ‘SOLI p “Sold g pur ‘g ‘p eee eee e een ene “ see eeenes ce Here eeenegTgg gE ‘samoy PF T09FV 0-G wee eee ce weenee weeeee 13 teense (73 sen nee “ec nee eee *10}BOTPUT 6L IT OT “ONT 99 004 U HORN fo sornuendy Sure Surureyzuoy suorynyog ut ssdq snuryoy uodn sioyeorpuy Jo yoy 047 Surmoyg 146 (7 poolg) TITA 4eL, 147 the Development of Hggs of Place and Echinus. 1905. ] “HOTFVUTUNCKS O}7V Uoyorq AToJwUNZAoFUN oQog Tf coded uo syanue teoury Aq Arwgiqre sqyUOTOMSTOT SS S89] «| -worout “Ey “ prog, te « as iss ee poyetoued poqra -op pus prog | -ouoseq “UOHVUTTUVXS to97v oyNSTUA Aq poppe UTPRUTLOF Fo quNoUTR [Peg 4 isi ay ce ce OL0FOq, 8B OURS "g pourtoy -ATprq [BLOAOG “ iff ce “ ‘tj cc Surpoooad uvyy —-ST[90 orour FT “TT tee a “Sonat W ‘WT ‘MOISIATp “BoaryT surpoooad uy} poeprarp S80] S199 ZT “seaNAOWL FT p pue ‘g ‘g “Sonat Ato A “Sout Axoa ‘oxour % “p Z Moy B “Som g ‘Kav dn Suryvoaq pus “Sot ‘Z [@t9A08 “10A0 pur p ATASoUR ‘% Moy @ “DOIIT ter eeenee 8 & pus ‘p 0.6 oT 8-0 v0 17 “ (tj ce «cc uropuyyyd -[ouoTy 8T LY 9T ST ¥L €T [Nov. 14, Mr. E. Whitley. fect of Acid, etc., on 148 “B[NASE[G OY} JO eoMETEFUMOIIO 9} UT poyUNOd s][e0 Jo LequINU eat MUMNJOO sIYy UI sLOqUINNT » eLOJOG SB oueG auo WNq ASB] SB oUILE 4svy UByy 10¥j0q° pvep ‘sesvys colp1ve 10 ao py “ 6c peop ‘amy, dn weyxorq 10 poyuemsesuyy p8ep ‘odes Lolptee 10 wMzoyy “ce iis “ce ce ce ce prop ‘apnayses Ajxeo [BUOIsvo00 pues apyselg [AED UNIS (a peep ‘xe [N.tour amos “padopeaop AT100g pvop ‘efayses Surouemtut0g peo xayses Ap1ve outos ‘apse peed dn suryevorq ‘s[jeo epSurg na800 T 118 Ajxvou ynq ‘xepnseum A104 QT pur ‘g ‘ge ‘g ApAsour ‘ee~aour ADIT Sc iy OL ce (79 Hee e meee weer enee ZI ce ce trreeeerresesces Or Comaserq ATER “ IeMPaIIt OT ‘ssoy pues vlna dn ueyorq pue weseaty ‘resseeeees TEMBOAIL ‘SSaT PUB ST[99 OT ““peumtog yok apngseyq ou “rep seary “-* Jurpeooid ueyy rosary ATyqSTS wee if OC gO SOR BO al DOS a UC UNIO a1-4 ‘elmgse[q. “ 97 “eamgaseyq ATrve pus alMAoyy “ : dn ueyorg aveiulole(e;s\e/xis)niujels(a\o[sielvisle/aiclalele late lore 9% ‘elngyselq, ““"+ negouUBApe aToUL ‘afngselq ATTeq “* g)T ‘eyseyq ATrve pues emmaoyy “SUTUL OF “SI FZ ‘SUITE OG ‘SI ZT LOFM-VIS °0°O OOS UL HOM JO sorqtquen’y SurArwA Suruyezuod suorynyjog ur sosy snuryory uo s10yeorpuy Jo yooHT— YT qe, rireeenecresreereeteeteeee 9 ge QUTEG aponoodaHoqoHONE - eT sv oureg “"" g sB oureg Pee e ee eneeneoee P pue @ qwepaseta pue 4 Pp ojoyd -ULOUI euLos pus *Z outos ‘Tt Auepy SUOISTATP LoY}.INF ojoTdunoout pus ‘FZ POC e ee eee eee ee sees eer ese eeeseeeas P Moy @ % “'"* SUOISIAIP TvpNserIt Moz B ‘8,7 teeeeeeseees Gn suryveaq pus remMsea1] seeneees “""G SB OULBG noeHCN OOOH Fe} Coons UWOISTATp xrelMsor11 Ve paapnaccaG090q0000 (4 eje[duto0our MOT B % Hence eee eee eee eeees eee eee eee eeeeeeees G7 940[dut0out eum0s % Pee nee bee eee e eens nsesesee PeprAlp suo Ny C}G IEA syjeo Moz ATOA ‘MOISTATP Iep~ndoaryT Pee eee eee ee eee nes (73 i13 fe enenee ce ce Pp pus ¢g [euorseo00 ‘Zz 09003011 o0aGDap oC OSS TOBEDEGHING SONIA GHD: r= eee neeereee ‘SUIT OG “SAY fF 109TW (Z pooug) en 0900 OI od <0 eo woo00 ANOTI9 Ona Fy g HOO © S SOHANDM FT 0 OF ‘HON 01/N jo yunoury Soo iy ce Peewee eeeee ce OOOO OOOO) (13 NOOO OOOO) “ weet eee eeeenee ce Peete eee eseene Ce see wees eecees cc DOO i iy ce Dee eee ent eeence *10JVOIPUT SNOAID © Fr DAD “ON 149 ae chinus. and g vce the Development of Eggs of Pla 1905. | WOTSIAyp qv sqydureqqe = aepnSeaat eulos 4nq 9g[suIs [RV Ajavou ‘apniout Ay«ve SS FWs peouvape ssey 4sB] UBTY peouvape ssoyT “B_N1OUL Ayara 8 ‘9 ‘p ‘Z ‘soySutg eTqequnod [Te ATrvou ‘encour ATIva 0} seTsUIg qsvy] WeYY Ivpnsorat aLOUL PUT PIdDUBAPB sso] BMISVG Moy vB ‘amniou zg ‘eTqeyunoo & BlMyseyq Suyetouesaq |*''''' pedopeaep [Jom JoN | MeF B soeeng pur ZT Apso podoyaaop ATVYsTIs AUS “SI FIT “say, TL TBTNZAIT ‘SUOTSTATP | AyJavo qso ‘sopsurs 488] UY} SO[SUIS SLOW TIS poyuetisesun [14s peyueMIsSesuN UIvUlet | TIS MoF B astncout Ape gy gse] Wary Lortve ATYSTTS @[nysv[q Surdopssep pue apniz0yy & qnoqge ‘eto MEF VW | pue g fF avpnsomtt qsor ‘2 Sue AToo.rvos ‘opsuts G B[NAOW LepNsotat | saa ‘s.g puokeq gou St qsox ‘*xRTNLOUL -ATpeq % ‘apnyse[q ON jenbe qnoqe a[n4se[q pure a#pnsou ysea ‘Z AU Tt ‘emiou | ‘aepyseyq "SULUL QE “SAU Lv “sat 61 (‘¢ poorg) “IOYVM-BOG 9'D YYZ OF B8VQ Yovs ul poppe ‘stQoUEIG SuTdiVA UI WepeyyYyd-jousyg Jo UoTNTOY oIPOYOoTY “yWeED Jog €.Q JO sHdq snuUTYIy Jo yuowdopose”y] uo yoyy— X e[qey jo yunowy ae GZ sorduoour Moz v ‘8 T ¥-0 Tg * &-0 0& poouvapr ssoT G0 66 ‘oqa & ‘e 4 poulcoz TV[NGOATI ouos ‘fp jetoass ‘°Z CL.0 83 t |p pur g ose “Z ATsoyL 1-0 LG TOTp1va 4sot ® pue ¢ ojotduioour = | outos ‘gz oqepdutoo otwog CO. 0 9% 0°0 ‘SUIT OG ‘SIT, eee ‘ON LXXVII.—B. VOL. 150 The Mammalhan Cerebral Cortex, with Special Reference to its Comparative Histology. I. Order Insectivora.—Preliminary Communication. By Grorce A. Watson, M.B., C.M. Edin, (From the Pathological Laboratory of the London County Asylums, Claybury.) (Communicated by Dr. F. W. Mott, F.R.S. Received July 28,—Read December 14, 1905.) The results and conclusions brought forward in this paper form a portion of the outcome of an extensive investigation dealing with the cortex cerebri in various orders of mammals. The work has special reference to the neopallium only, and has for its prime purpose an endeavour to shed some further light upon the functional significance of the cerebral cortical lamination. Animals Hxamined and Methods of Study. The brains of the animals belonging to this order examined are :— 1. The Mole (Zalpa Europea). 2. The Shrew (Sores vulgaris). 35. The Hedgehog (Hrinaceus Huwropeus). The cerebral cortex has been examined by means of complete series of sections cut in almost every possible direction and stained by one or other modification of the Nissl method. As part of the method of study throughout the entire investigation, the natural habits of the animals examined, and their educability, as far as facts relating to the latter are available, have been considered when attempting to correlate structure and function. Macroscopic Appearances and Microscopic Furrows. All are almost smooth highly macrosmatic brains. That of the Hedgehog is one of the simplest mammalian brains. It presents in addition to the rhinal fissure a short presylvian furrow; the latter is not found macros- copically in the Mole and Shrew, but is seen on microscopic examination of sections. By this method also, in the Mole only, two shallow more or less longitudinal curved furrows can be traced on the dorso-lateral aspect of the hemisphere, which appear to represent foreshadowings of the corono-lateral and supra-sylvian sulci. Signs of greater differentiation of the neopallium of the Mole, as compared with the Hedgehog especially, are further exhibited on microscopic exami- nation of the structure of the cortex. The optic nerves in the Mole and The Mammalian Cerebral Cortex. 151 Shrew are reduced to small threads. In the Hedgehog these nerves are considerably larger. In all three the fifth nerves are relatively very large. The Lamination of the Neopallium. General Remarks.—The classification of the cortical layers adopted by the writer is that introduced by J. Shaw Bolton. The latter considers that the human cerebral cortex is constructed upon a five-layered type—viz., {, Molecular; II, Pyramidal; III, Granular; IV, Inner line of Baillarger: V, Polymorphic, Of these only three are primarily cell layers—viz., the pyramidal, granular, and polymorphic, Layers I and IV being primarily fibre layers, although containing nerve cells—the cells of Cajal in Layer I, and the Betz cells (psycho-motor region) or solitary cells of Meynert (other regions) in Layer IV. The outstanding features of this classification are: (1) The recognition of the granular layer as separating the true pyramidal /ayer above from the more or less pyramidal shaped cells which may be found below this layer, for the cells of Layer IV are not “ pyramidal” cells at all, the Betz cells in the psycho-motor area constituting “the origin of the important tract for skilled voluntary movement,” whilst the solitary cells of Meynert in other regions “probably possess a somewhat analogous function.” (2) The con- sideration of the pyramidal layer as forming one layer developmentally and functionally. Bolton, as the result of his studies of the development of the human cerebral cortical layers, and of their depth in the normal individual as well as in various degrees of amentia and dementia, has come to the following conclusions as to the functions of the three primary cell layers. The pyramidal layer “subserves the psychic or associational functions of the cerebrum.’ The granule layer “ probably subserves the reception or immediate transformation of afferent impressions, whether from the sense organs or from other parts of the cerebrum,” whilst the fifth, or polymorphic layer, “ probably subserves the lower voluntary functions of the animal economy.” When dealing with the mammalian cortex generally one or two further explanations are necessary. The term “granular” is used in a wide generic sense and as indicative of a certain cortical layer rather than of the cell constituents of this layer, which latter, in an adult animal, may take the form of angular, quadrilateral, stellate or even small pyramidal-shaped cells, or a mixture of these elements. In some regions of the cortex in certain animals these elements of the granular layer may be scattered and comparatively few in number; yet their recognition is of importance, for such provides the means by which the lower limits of the true pyramidal layer may be determined. Owing M 2 152 Mr. G. A. Watson. [July 28, to the difficulty experienced in accurately separating the fourth and fifth layers (which tend to intermingle) in the cortex of some animals, the writer prefers to speak of these layers together as infra-granular. It is also proposed for the sake of definiteness to term the true pyramidal layer (‘.c., Layer IT) the supra-granular layer. Areas in the Neopallium. The appearances of the cortex in the Mole and Shrew being very similar, the following, concerning the cortex of the Mole, may be taken as applying also to the Shrew excepting when the latter is specially mentioned :-— 1. Dorso-lateral Surface-—In the Mole this region presents two main and distinct types of cortical structure, with certain areas of comparatively undifferentiated cortex (fig. 1). x x (0) (0) °° [0) ®2e@o0® ®FPOX2POO GO (0) \ \ \ \ { t 1 1 06090999 ©00O0000 06 Fic. 1.—Dorsv-lateral view of the Right Hemisphere of the Mole. Left-hand figure. x Ayvea 1, “Motor”; x thesame, but less characteristic ; © Area 2, General Sensory ; © the same, but less characteristic ; 5 Archipallium. The anterior, lateral, and posterior areas of undifferentiated cortex are left blank excepting the portion of neopallium represented as shaded, which is thinner than the remainder. Right-hand figure. a, vhinal fissure ; b, c, and d, probable representatives respectively of the corono-lateral, supra sylvian, and presylvian sulci; 6, c, and d, vary much in individual distinctness in different hemispheres. The figure is a composite one. Area I: Motor.—This extends antero-posteriorly from a short distance 1905. | The Mammalian Cerebral Cortex. 153 behind the anterior pole to about the posterior quarter of the hemisphere, and laterally from the dorso-mesial margin (or more or less close to this anteriorly and overlapping this posteriorly), to about half-way between the dorso-mesial edge and the rhinal fissure, the lateral limits varying somewhat at different points (figs. 1 and 2). Area IT: General Sensory.—This occupies an extensive region lateral to Area I, but does not reach as far as the rhinal fissure, being separated from the latter by a zone of undifferentiated cortex (fig. 1). The features of Area I are less characteristic mesially, anteriorly and posteriorly, and of Avea II anteriorly and posteriorly, These two fields appear in every way to be the best developed areas of the neopallium and to be amongst the oldest. phylogenetically. Owing to their different histological appearances, Areas I and II in the Mole can be readily delimited. From the relatively greater numbers and prominence of the large cells (homologues of Betz cells) in Layer IV, it is coneluded that Area I possesses especially motor attributes, and that on account of the ereater development of the granular layer throughout Area II, the latter “Motor” ~ e of vay x7 5 x? S > NWN ZSTENORS KKK NJ ONG na x XX x na v7 * SAS tn Sake eee yee Tens, gems Bie _ , Fig. 3.—Postero-mesial aspect of the Left Hemisphere of the Hedgehog, for comparison. with fig. 2 (Mole). Mole) and the cortex behind and below this, including archipallium on the one hand, and the posterior part of the corpus callosum and hippo- campus on the other. The area is not of the same shape in the Mole as in the Hedgehog, owing to the hemisphere being moulded in a different way in the two animals. This region, although characterised throughout by the well-marked granular layer referred to, presents certain differences of detail, owing to which it has been separated into three divisions in each animal. The anterior and superior portion (Division 1) is of about the same relative size in the Mole as in the Hedgehog. In both there are good granular and infra-granular layers, and in the Hedgehog a shallow but definite supra-granular layer. The middle portion (Division 2) is relatively 156 Mr. G. A. Watson. [July 28, larger in the Mole than in the Hedgehog, but in both, comparatively to Divisions 1 and 3, it presents unspecialised features. The inferior portion (Division 3) is relatively not.only considerably larger in the Hedgehog than in the Mole, but in the former it is a more histologically distinct field, showing not only a well-formed “granular” layer (the cellular elements in which, however, are mostly angular or small pyramidal in shape), but a comparatively good infra-granular and a definite, though shallow, supra- granular layer. In the Shrew this division is rather better developed than in the Mole. Owing to the presence throughout this region of such a deep and definite granular layer, it is concluded that this field is sensory in function, and regarding it the following suggestions are made :— (a) The area is much too large to be concerned only with the cortical distribution of the optic nerves, which are relatively so minute in these animals, especially in the Mole. (6) The inferior portion (Division 3) only is visual. In the Mole this portion is a mere vestige; in the Hedgehog it is better developed and relatively larger, having apparently extended upwards somewhat and encroached upon the area of unspecialised granular cortex (Division 2) as compared with the Mole. (c) The middle and superior portions (Divisions 2 and 1) may correspond to the large infra-calcarine area of certain relatively higher mammals (¢.¢., Ungulata and Carnivora), in which, owing to the greater development of the visual faculty, the inferior portion (Division 3, visual) has, so to speak, extended upwards, backwards, outwards, and forwards so as to overlie the middle and superior portions, and has become the calearine region. (d) The relatively well-developed superior and anterior portion (Division 1) of the two specialised divisions in both Mole and Hedgehog may be concerned with the cortical distribution of the fifth sensory nerve. On account of the importance of the fifth sensory nerve as an avenue of information to these mammals through snout or vibrissee touch, or both, and, in view of the large size in them of the fifth nerve, it seems probable that the sensory portion of this nerve should have a very special cortical representation. The Cerebral Cortical Layers. (Neopallium.) Although the total depth of the cortex in the best developed regions is different in the Mole, Shrew, and Hedgehog, the relative depth of the separate layers, supra- and infra-granular particularly, appears to be about the same in all. The following micrometric measurements of the cortical layers in three areas of the Mole’s cortex, which have been kindly furnished 1905. | The Mammalian Cerebral Cortex. 157 by Dr. J. S. Bolton, may be taken as fairly typical also of the relative 70 4 hours, ammonium sulphide, glycerine. x 590. Fig. 16.—From the spinal cord of frog, illustrating a typical Frommann striation, as revealed by the mercurous nitrate method. Hg,(NO;), =. 30 minutes, ammonium sulphide, glycerine. x 590. Macallum and Menten. Roy. Soc. Proc., B. vol. 77, Plate 2. Pic 2: Fic, 3. TEs h Sg rt < ‘ f 2 4 be E 1 < - ; ro rr 3 Seon “| E aoe ess. «| E | 2 i £ wenn ees = ¢ > { : Abr & i = j } s 7 ®, rt Fi = % iat tr = g f ¥ ® i } m2 £ b 2 LEESEL ILIAD a re aes terse, : Lhe eae ’ Fit ; ¢ . i Fie. 4. : es a b ier t SOU eyrtce gyn wh oe ruuee esa StH MUSE RU ne UOATIC AUER: \ Fig. 6. b c geen d SURYA Ks FESS Mesa Te apey rect verses P Macallum and Menten. Roy. Soc. Proc., B. vol. 77, Plate 3. Fic. 9. ; Fic. 12. a b a ; res 10: 5M Ve ed ue a b (ee aeons pute te USF HHA fara fab) # ¥ Rg: ime pimpin, i S . oa ef ! Wek eee] SNe = , 5 es Ste ‘ 7 it Rec nae fF et Pe | 2 f J ay Sees wee < ; 22 ae} Ss Fr cate a. en) = rf eee Se 3 —= f-4 hehe.” “ =: é A %, ES Be a % Sek : \ ity Be: ge colgee < i a RPS oes ° he = i : = neal 2 hoy a aie Ke y > 3 es 4 tof ne oy! = peas Ten See Sa Be =) Fis oe Ff -! 74 is sta ani = : eM ta } [ed met ae ¢ % ) Fs sen =e : : aS, = = J 5 z % 2 ' ; 5 Ne: = Flava E he ece t as Reegateys : (: iy 2 ita: ay E Le ig = Cis 3 5 A ' = ‘ ~ Se ee a ae ae Macallum and Menten. Bie. 17. Pre 18. , : WS ve # be. a a . +» ea a as , Y i +: rf he Pr ¥ y ’ \ i <%, ta Roy. Soc. Proc., B. vol. 77, Plate 4. Fig. 19. i) Hire? 20: th WN i AMA Se se wee +% eA me Rice. 2 a) 1905.] Distribution of Chlorides in Nerve Cells and Fibres. 193 PLATE 4. Fia. 17.—From the sciatic of frog, showing the occurrence of chlorides in the imbrica- tions of Lanterman. x 590. Fic. 18.—From the sciatic of frog, illustrating a distribution of chlorides in the medullary sheath sometimes found. Hg,(NO,), e 30 minutes, ammonium sulphide, glycerine. x 590. Fic. 19.—From the sciatic of frog, illustrating the occurrence of chlorides in the imbrications of Lanterman and in the axon. Hg,(NO,), = 2 hours, ammonium sulphide, glycerine. x 70. Fig. 20.—Portion of end of capillary tube filled with albumen. AgNO, = +15 per cent. HNO,. The reagent has diffused in the direction of the arrow. x 105. Fig. 21.—Portion of another capillary tube similarly treated. x 105. Fic. 22.—Portion of a capillary tube filled with albumen and treated as in the case of fig. 20. x 590. Fie. 23.—Portion of a capillary tube filled with albumen, to which sodium chloride was added to bring the strength of the salt in solution up to 1°73 per cent. Treated as in the case of fig. 20. x 70. Fic. 24.—Portion of a capillary tube filled with gelatine containing 0:288 per cent. of sodium chloride, illustrating irregularities in the distribution of the silver precipitate. Treated as in the case of fig. 20. x 90. Fies. 25 and 26.—Portions of capillary tubes filled with commercial gelatine and treated as in the case of fig. 20, illustrating peculiarities in the precipitation of the silver chloride formed. Fig. 25 x 590, fig. 26 x 105. 194 On the Possibility of Determining the Presence or Absence of Tubercular Infection by the Exanunation of a Patient's Blood ~ and Tissue Fluids. By A. E. Wricut, M.D., sometime Professor of Pathology, Army Medical School, Netley ; Pathologist to St. Mary’s Hospital, London, W.; and Staff-Surgeon 8. T. Rem, R.N. (From the Pathological Laboratory, St Mary’s Hospital, London, W.) (Communicated by Sir John Burdon Sanderson, Bart., F.R.S. Received October 21,—Read November 23, 1905.) In the present communication we propose (a) to set forth certain con- clusions arrived at after the study of the tuberculo-opsonic power of the blood in a very considerable number of tubercular patients; (0) to show that we have in the measurement of the tuberculo-opsonic power of the blood and tissue fluids a method which may be exploited in the diagnosis of tubercular infection. Technique Employed—tThe technique employed by us in the measurement of the tuberculo-opsonic power of the blood was essentially that described by one of us in conjunction with Douglas.* In each case the white corpuscles required for the tests were derived from blood from the finger received into a solution of 0°5 per cent. citrate of soda in 0°85 NaCl, and rewashed after centrifugalisation in a considerable volume of 0°85 NaCl, and then again centrifugalised. Of the “blood-cream,” obtained by skimming off the upper layer of the corpuscular sediment, one portion was in each case mixed in a capillary tube with one volume of serum and one volume of a suspension of tubercle bacilli which had been centrifugalised in such a manner as to free it from bacillary clumps. After incubation at 37° for 15 to 20 minutes films were made on slides prepared with emery paper.t These films were, after fixture in saturated corrosive sublimate, stained with boiling carbol-fuchsin, decolourised with 2 per cent. sulphuric acid, and counter-stained with methylene blue after washing in 1 in 1000 sodium carbonate. The standard of comparison employed was obtained by mixing in each case the same “blood cream” and tubercle suspension with “pooled normal serum.” * ‘Roy. Soc. Proc.,’ vol. 72. + Wright, ‘ Lancet,’ July 9, 1904. { While in this research pooled serum was employed, in order to provide against any chance variation of our bloods under the physical strain entailed by the work, it is to be - noted that the observations of Urwick, conducted in this laboratory, and the more exten- On Tubercular Infection in a Patient's Blood, ete. 195 This “pooled serum” was obtained by mixing equal volumes of the sera of six to eight healthy students or laboratory workers. We have found that the opsonic power of such a “ pooled serum” corresponds to the arithmetical mean of the opsonic indices of its component sera. Classification of Tubercular Cases into Strictly Localised Cases, and Cases which are associated with Constitutional Disturbance. Cases of tubercular infection distribute themselves in a natural manner under two headings. Into one category would fall the patients who are the subjects of a strictly localised infection unaccompanied by anything in the nature of constitutional disturbance. Cases where the infection is limited to one or more lymphatic glands; further, most cases of lupus, most cases of tubercular abscess in the subcutaneous tissue, tubercular affections of the joints, and, lastly, many stationary or only slowly progressing cases of tubercular phthisis, fall into this category. Into another category would fall patients who are suffering from more generalised tubercular infections associated with constitutional disturbance. This group consists in large part of cases of pyrexial pulmonary tuberculosis. With these may be classed certain other cases of extensive or widely dis- seminated tuberculosis. Data with regard to the Tuberculo-opsonic Power in Cases of Strictly Localised Tuberculosis, The opsonic index is here low and uniformly low—in exceptional cases as low as one-sixth of the normal. Our findings in a series of cases of strictly localised tubercular infection are appended in tabular form below. sive series of investigations carried out by Bulloch at the London Hospital (‘Medico- Chirurg. Soc. Proc.,’ 1905), and Lawson and Stewart at the Banchory Sanatorium (Joc. cit.), have conclusively shown—(a) That the tuberculo-opsonic power of the blood does not in health range below 0'9 or above 1'1; and (6) that the bloods of A. E. W., S. R. D., and others which have hitherto in this laboratory furnished a standard of comparison, are, from the point of view of their tuberculo-opsonic power, typically normal bloods. 196 Dr. A. E. Wright and Staff-Surgeon 8. T. Reid. [Oct. 21, Table I—Showing the Tuberculo-opsonic Index in a Series of Cases of strictly localised Tuberculosis. Initials or, as the Tapert ; Serial No. case may be, Nature and seat of the infection. baat doc ae initial of patients. eee 1 J. R. Muberelelobitestisn. sr s..cccs+ sec fe eeseee- cesses 0°65 2 A. Caries, lower end of femur...............++- 0:7 3 C.8. Tubercular ulceration, dorsum of hand 0°86 4 A. B. Hs iritis. 0°51 5 B. C. _ fEllgnEGlS coooacnacdoo0aR—s590000020 0°4 6 E. M. op ulceration of legs. 13 years’ GIPENHIOSS _ ieagoseaancasacoocded ~O aly 7 H. W. By glands (neck) ............:0008 0 82 8 D. W. i “6 eM Lek id: 0°64 9 8. - ulceration of legs ............ 0-49 10 L. B. PF glands (abdominal)............ 0:13 11 iB: é AGM OV Ae eeeteaccss ies aceet sneer 0°75 12 M. H. . GYUDON) ocisosadsosdabongdboonde0 260 0°85 13 D. B. iy glands. Extirpated and re- IOS! sooso9opdvan50o050n|A 0.85 14 M. O. #, 3) glandsteespeesceedettseeaeaciege 0°6 15 W. Psoas/abscessy didi. tlhtane ts dscesesseeeteecters 0°75 16 C. H. - sialyl dsscbanenemssuseeneies swyesectielses 0°65 17 W. Tubercular glands (neck). 18 months’ duration seeecse a. seses sete 0°47 18 C. sn. SWIRL Gem score ae: 0-85 19 U. RB. 3 glands (neck) .............0000 0-7 20 A. H. . glands. Hxtirpated and re- BPPCALEA Je.o.scaceeneeesnene 0°54 21 Bs ” THUKO OTS — saonocnopancconcndeda 06 22 R. a abscesses and glands ......... 0°6 23 E. TDW SIS) Saqqa6sd00s00cboncoqcRan6EcdoNd00ed000600 06 24 R. Tubercle of testes and bladder ............ 0-72 25 1 Tubercular peritonitis ...................4. 0°7 26 H. 55 caries of fibula ............... 0°6 27 C. Tubercle of kidney ..............sccseeneeneee 0°88 28 W. Tubercular disease of knee ............... 0°6 29 4, % glands (neck) recurrence ... 0-66 30 8. 5 ulcer Of fOOt ...2......20..0008 0-49 31 C. 4 disease of knee .............+ 0°75 Data with regard to the Tuberculo-opsonic Power of the Blood in Cases of Tuberculosis associated with Constitutional Disturbance. In the cases here in question the opsonic index of the blood is continually varying. The range of its fluctuation is from considerably under the normal to twice or more the normal height. Striking examples of the variation of the opsonic index in connection with acute tubercular phthisis are furnished in the paper of our fellow-worker, R. H. Urwick, already referred to.* The following are instances of similar variation occurring in the subjects of other forms of tubercular infection :— * ‘British Medical Journal,’ July 22, 1905. 1905.] On Tubercular Infection in a Patient's Blood, ete. 197 Example 1—Child with Tubercular Caries of the Fibula, associated with Constitutional Disturbance. Dates of blood Tuberculo- examinations. opsonic index. 11.9.05 1:45 14.9.05 17a 19.9.05 13 28.9.05 0:98 30.9.05 Operation, fibula scraped. 2.10.05 Tis 3.10.05 1-13 10.10.05 13 Example 2.—Adult Patient with Tubercular Caries of the Spine and Constitutional Disturbance. Date of blood | Tuberculo-opsonic examination. index. Remarks. meee Coie: os Temperature disturbance and pain associated 22.5.05........ 13 with development of abscess. Ss OO ccwaccas’ 1°0 Temperature returns to normal in association 24.5.05..c..0e8 0°8 with spontaneous discharge of abscess. Example 3.—Adult Patient with extensive Psoas Abscess and Generalisation of Tubercle. Case has since terminated fatally. Date of blood Tuberculo-opsonic examinations. index. 8.2.05 2 9.2.05 24 11.2.05 0°6 Suggested Interpretation of the Different Findings in these two Categories of Cases. The explanation of the difference in the condition of the blood in these two contrasted categories of cases is probably the following: The condition of low opsonie power which is associated with strictly localised tuberculosis is almost certainly a condition which has preceded and has furnished the oppor- tunity for infection. The fact that the opsonic index continues persistently *A rise in the opsonic power similar to this here registered has been repeatedly observed by us in connection with the stirring up by surgical interference of tubercular foci. VOL. LXXVII.—B. Ip 198 Dr. A. E. Wright and Staff-Surgeon 8. T. Reid. [Oct. 21, low after infection has supervened, while it can invariably be raised by appropriate inoculation,* indicates that the machinery of immunisation with which the organism is furnished is not, under the conditions which obtain in strictly localised tubercular infections, spontaneously called into play. The constant fluctuation in the opsonic power of the blood in cases of active pulmonary tuberculosis and other active forms of tubercular infection furnishes—as we can hardly doubt—evidence of a periodic conveyance of tubercular elements into the blood ; and of a response to such stimulation on the part of the machinery for immunisation. The low opsonic indices registered in connection with active tuberculosis would in other words be “negative phases” such as supervene—as one of us has shown—upon the inoculation of all vaccines; the high opsonic indices would be “ positive phases,” such as normally succeed upon the negative phases just mentioned ; and the normal opsonic indices would correspond to periods of transition between negative and positive phases, or, as the case may be, to periods in which the blood is returning after a positive phase to the condition quo ante. The life of a patient with any really active form of tuberculosis would in conformity with this view be a life of alternating negative and positive phases: the favourable or unfavourable event of the infection being in each case determined by the adjustment or want of adjustment of the auto- inoculations (with respect to dosage and interspacing) with the particular patient’s capacity for immunising response. - Having now to a certain extent cleared the ground, we may pass on to consider the question of the diagnosis of tubercular infection by means of the measurement of the opsonic power of the blood. Exploitation of the data Summarised above as an aid in the Diagnosis of Tubercular Infection. Consideration will make clear that the data obtained by the measurement of the opsonic power in cases of doubtful diagnosis may, when adjudicated upon in the light of the data obtained in connection with undoubted cases of tuberculosis as given above, furnish material for admitting or rejecting the diagnosis of tubercular infection. We may formulate in connection with this matter the following propositions :— (1) Conclusions which can be arrived at when we have at disposal the results of a series of measurements. (a) Where a series of measurements of the opsonic power of the blood reveals a * Exactly the same statements hold true with regard to the staphylo-opsonic power in localised staphylococcus infections (furunculosis, sycosis, etc.). 1905.] On Tubercular Infection in a Patient's Blood, ete. 199 persistently low opsonic power with respect to the tubercle bacillus, it may be inferred, in the case where there is evidence of a localised bacterial infection which suggests tuberculosis, that the infection in question is tubercular in character. (b) Where repeated examination reveals a persistently normal opsonic power with respect to the tubercle bacillus, the diagnosis of tubercle may with probability be excluded. Illustrative case: A. B.—Case diagnosed as tubercular cystitis on the evidence of pus in the urine, of the cystoscopic appearances and general disturbance of health. The measurement of the tuberculo-opsonic power of the blood yielded the following results :— Date of blood Tuberculo-opsonic examination. index. 2.3.05 0-98 14.4.05 0:99 28.4.05 1 18.5.05 1 19.5.05 11 2.10.05 0:97 The inference that the cystitis and disturbance of health was not of tubercular origin was confirmed (a) by the fact that an extensive series of bacteriological examinations prolonged over many months revealed in every case the presence of proteus in large numbers, while the tubercle bacillus was never found, even when examined for by the inoscopic method of Jousset; (b) by the fact that the patient’s blood possessed, anterior to treatment with regard to the proteus, an agglutinating power which was three times higher than the normal; and (c) by the fact that very striking amelioration of the cystitis, and a complete return to health has been obtained as the result of the inoculation of a proteus vaccine. (c) Where there is revealed by a series of blood examinations a constantly Jluctuating opsonic index the presence of active tuberculosis may be inferred. C. D—A case of severe chronic urticaria of unknown etiology. The measurement of the tuberculo-opsonic power of the patient’s blood yielded the following results :— Date of blood §Tuberculo-opsonic examination. index. 20.5.05 13 26.8.05 13 16.6.05 0°86 20.6.05 1:27 200 Dr, A. E. Wright and Staff-Surgeon S. T. Reid. [Oct. 21, The inference drawn from these data that the patient was suffering from some active form of tuberculosis was confirmed (a) by the discovery by an independent observer of a lesion in the apex of one lung; (0) by the subsequent development of an abscess of an obviously tubercular character; and (c) by the marked improvement in health which has followed upon inoculation with tubercle vaccine. (2) Conclusions which may be arrived at where we have at disposal the result of an isolated blood examination. (a) Where an isolated blood examination reveals that the tuberculo-opsonic power of the blood is low, we may—according as we have evidence of a localised bacterial infection or of constitutional disturbance—infer with probability that we are dealing with tuberculosis—in the former case with a localised tubercular infection, in the latter with an active systemic infection. (b) Where an isolated blood examination reveals that the tuberculo-opsonic power of the blood is hiyh, we may infer that we have to deal with a systemic tuberculous infection which is active, or has recently been active. (c) Where the tuberculo-opsonie power is found normal, or nearly normal, while there are symptoms which suggest tuberculosis, we are not warranted, apart from the further test described below, in arriving at a positwe or a negative diagnosis. Discrimination of Tubercular Blood from Normal Blood by the ad of the Phagocytic Test Conducted with Serum which has been sulyected to a Temperature of 60° C. The further criterion to which reference was made in the preceding paragraph is the following :— When a serum is found to retain in any considerable measure, after it has been heated to 60° for 10 minutes, its power of inciting phagocytosis we may conclude that “ incitor elements”* have been elaborated in the organism either m response to auto-inoculations occurring spontaneously in the course of tubercular infection, or, as the case may be, under the artificial stumulus supplied by the inoculation of tubercle vaccine. A typical selection from the very extensive body of observations which furnishes the basis of the above statement is presented in Tables II and III. It will be seen from these tables that in practically every case where a reaction to tubercular infection may be assumed to have taken place, * The term “incitor elements” (Latin, incito, I urge forward, I hasten, I bring into rapid movement) is here employed in lieu of a more specific term, in order not to prejudge the mode of action of the element in the heated serum which promotes phagocytosis. The nature of the incitor element is considered in the next following communication. 1905.] On Tubercular Infection in a Patient's Blood, ete. 201 Table I1.—Showing that the Normal Serum, after it has been exposed to a Temperature of 60° C. for 10 minutes, no longer incites Phagocytosis.* Unheated serum. Heated serum. Serial i number| Derivation of the | Phagocytic count. Phagocytic count. of the eae (Number of Tuberculo- (viele oe Tuberculo- observa- bacteria ingested . bacteria ingested : tion divided b papi divided b Weise : y index. y index. number of leuco- number of leuco- cytes examined.) cytes examined.) 1 Healthy man ......... (104/40) = 2°6 | Takenas1 Caio} = 0°32 0°125 PA allie? Ua) Sel SanPep rane (96/40) = 2°4 1 8/40) = 0°2 0:08 3 Pooled serum of six (247/86) = 6°8 i 1.| (30/50) = 0°6 0:09 healthy men 4 | Healthy boy............ (250/39) = 6-4 Gt ae = 0 a 0:06 AN a ae a (214/30) — 7-0 1 | (19/40) = 0-4 0-06 6 Pooled serum of eight| (60/50) = 1:2 i 1} (2/20) = 0°71 0-08 normal men 7 Healthy man ......... (55/40) = 1°4 55 1} (0/40) =0-0 0:00 8 Pooled serum of six | (182/80) = 4°4' of 1} (8/30) = 071 0-1 healthy men evidence of that reaction can be obtained by conducting the phagocytic test with serum which has been heated to 60° C. for 10 minutes. The observations numbered 15 and 16 respectively have, it may be noted, been introduced into the table with the special design of showing the very simple nature of the investigation which is required for the diagnosis of tubercle in the case where that infection has called forth a reaction of immunisation. The following observations, which we owe to our fellow-worker Dr. G. W. Ross, bring out in an instructive manner the trustworthiness of the phagocytic test with heated serum as applied in this its simplest form :— Case 1.—Girl, et. six years, Tentatively Diagnosed Pulmonary Phthisis. Phagocytosis obtained with the serum, heated for 10 minutes to 60° C. and employed in a phagocytic mixture containing over 1 per cent. NaCl. The verdict of tubercular infection of the lung which was based on this was confirmed on post-mortem examination. * In order to avoid the fallacies associated with spontaneous phagocytosis (wde the next following communication) the observations which are recorded in this and in the subsequent table were in each case made by mixing the volume of the serum with one volume of corpuscles, washed in 0°85-per-cent. NaCl solution and one volume of tubercle bacilli suspended in a 1*5-per-cent. NaCl solution. In this manner a salt content of over 1-per-. cent. NaCl was achieved in the phagocytic mixture. 202 Dr. A. E. Wright and Staff-Surgeon 8. T. Reid. [Oct. 21, Table III.—Showing that an element which incites Phagocytosis is contained in the heated serum of patients who are the subjects of an active systemic tubercular infection, or who have been subjected to inocu- lations of a tubercle vaccine. The sera, like those which are in question in Table II, were in each case heated to 60° C. for 10 minutes. Serial number of obser- vation. Ne) {e0) “TO Or iB Co bo H ra f=) a e 12 13 14 15 16 Nature of infection. Tubercular caries of hip Tubercular phthisis ... oP) 5; pe Tubercular peritonitis » OO Phthisis and tuber- cular glands Tubercular caries of hip Tubercular abscess of kidney : Lupus under treat- ment by inoculation of tubercle vaccine Tubercular ulcer of leg under treatment by inoculation of tubercle vaccine Tubercle of kidney under treatment by inoculation of tubercle vaccine Tubercular glands and abscess under treatment by inocu- lation of tubercle vaccine Tubercular cystitis under treatment by inoculation of tubercle vaccine IPhthisisiesesecase see: Unheated serum. Tuberculo- opsonic : index Phagocytic | (determined (Number of By bee bacteria Haaeeetic ingested divided | P nae! by number of : if leucocytes bes a examined.) Se with pooled blood of healthy men). = 15 (125/20) =6:2) 14 .| (152/80) = 5-0; 12 (98/30) = 3-2 10 (144/30) = 4:8 14 | (142/30) = 4°7 1°4 (113/40) = 2°8 et (110/30) = 3°6 1-0 _ Wy (34/10) =3-4| 0-7 (249/40) = 6-2 1 i) (68/40) = 1-7 15 (59/40) = 1°5 1°4 (97/50) = 2-0 = Heated serum. Tuberculo- opsonic index Phagocytic | (determined count. by compari- (Number of son of bacteria phagocytic ingested divided count by number of with that leucocytes obtained with examined.) unheated pooled blood of normal men). = 0-4 (113/30) = 3-7 0°8 (96/30) = 3-2 0-72 20/65) =0-°3| O-1 (103/30) = 3-4 10 (16/50) = 0°3 0-09 (79/50) =1°6| 0-6 (85/30) =2°8) 08 (26/30) =0'8| 0-4 (49/30) = 0:33 (149/40) = 3:7 0-7 | (77/40) = 1°9 1cY/ (36/40) =0-9| 0°8 (43/30) = 1-4 = | (26/80) = — (9/5) = = 1905.| On Tubercular Infection in a Patient's Blood, etc. 208 Case 2.—Man, et. 41, Tentative Diagnosis, Pleuwrisy due to Malignant Disease, or Tubercular Pleurisy. No phagocytosis obtained with the serum, heated for 10 minutes to 60° C. and employed in a phagocytic mixture containing over 1 per cent. NaCl. The verdict of pleurisy due to malignant disease, which was based on this, was confirmed on post-mortem examination. Case 3.—Case Tentutiely Diagnosed Miliary Tuberculosis or Malignant Endocarditis. No phagocytosis obtained with the serum, heated for 10 minutes to 60° C. and employed in a phagocytic mixture containing over 1 per cent. NaCl. The verdict of malignant endocarditis which was based on this was confirmed on post-mortem examination. Observation 4.—Case Diagnosed Miliary Tuberculosis. No phagocytosis obtained with the serum, heated for 10 minutes to 60° C. and employed in a phagocytic mixture containing over 1 per cent. of NaCl. The post-mortem examination revealed a complete absence of tubercular lesions and a healing typhoid ulcer* in the ileum. On two other Methods by which a Diagnosis of Tubercular Infection can be arrwed at or Hxcluded. In addition to the methods which have been already considered, there are two further methods which can be exploited in connection with the diagnosis of tubercular infection. The first of these is applicable where we desire to supplement the often ambiguous data furnished by the clinical symptoms in the case of inoculations of tuberculin undertaken for diagnostic purposes. The second is applicable where we can obtain, in addition to the patient’s blood, also lymph, or, as the case may be, pus from the seat of infection. Diagnosis of Tubercular Infection by the Aid of Measurements of the Opsonic Power carried out in Connection with the Inoculation of Tuberculin for Diagnostic Purposes. Already, three years ago,f in connection with a paper on staphylococcus inoculations as applied to the treatment of acne, furunculosis, and sycosis, attention was directed by one of us to the close analogy between the tubercular reaction of Koch and the local inflammation and _ general constitutional disturbance which may supervene when a patient whose * A negative Durham-Gruber reaction had been obtained in this case. + ‘Lancet,’ March 29, 1902. 204 Dr. A. E. Wright and Staff-Surgeon S. T. Reid. [Oct. 21, tissues are extensively invaded by the staphylococcus is inoculated with the corresponding vaccine in such a manner as to develop a pronounced negative phase. The association of a negative phase with a reaction similar to that conveniently spoken of as the tuberculin reaction, suggested to us the propriety of enquiring whether the true tuberculin reaction, as seen after the injection of Koch’s old tuberculin into a tubercular patient, was also associated with a negative phase. The opportunities for investigating the question which have presented themselves have not yet been sufficiently numerous to allow of our formulat- ing a final answer to this question. The observations which are set forth below seem to us to suggest that the development of a negative phase, with a dose of tuberculin smaller than that which would produce this result in a healthy patient, may prove to be an index of tubercular infection. Such a conclusion would be in harmony with our experience in connection with the therapeutic inoculation of tubercle vaccine (new tuberculin). We find in this connection that the negative phase supervenes upon very much smaller doses and persists much longer in the case where the patient is the subject of extensive infection than in the contrary case. Observation 1.—Case diagnosed, Tubercular choroiditis. Date. | Tuberculo-opsonic index. Clinical data. 26.4.0B...s..00c0. | 0:9 5 milligrammes old tuberculin inoculated. 0-29 Some constitutional re- 0°95 action, ¢. 100° F. Observation 2.—Case diagnosed, Lupus erythematosus. Date. | Tuberculo-opsonic index. Clinical data. TOO, ee | 0-73 13.1.05............ 0°85 No rise of temperature OD eeeer cece 1°6 or constitutional or A2GERODweaseasences 0°5 local reaction. 1905.] On Tubercular Infection in a Patient's Blood, etc. 205 Observation 3.—Case diagnosed, Lupus erythematosus. Date. Tuberculo-opsonic index. Clinical data. O34: Operarsgess ses 0 “66 Tnoculation of 5 milligrammes of old tuberculin. | | LALO: recksssnes 0-7 | Quite insignificant con- NDA Ob secccceses 1°2 | stitutional disturbance. TAA OD... cwacecices 0°85 Observation 4.—Case diagnosed as Lupus vulgaris. Date. Tuberculo-opsonic index. Clinical data. Inoculation of 5 milligrammes of old tuberculin. PAO bi oenccncases: iL ct Quite insignificant con- 12.4.05............ 1:0 | stitutional reaction. 14.4.05.........00. 1°0 | Observation 5.—Lupus, patient had been treated for many months by therapeutic inoculations of tubercle vaccine. Date. Tuberculo-opsonic index. Clinical data. PAP O55 kad, | 1°4 Inoculation of 30 milligrammes of tuberculin. QDs ODT eee rece | 0°34 Severe constitutional and ZOM OSS cercccsasel 21 local reaction, ¢. of PELOo ee 17 103° F. Diagnosis of Tubercular Infection by the Comparison of the Opsonic Power aj ' the Patient's Blood with the Tuberculo-opsonic Power of the Fluids Derived from the Focus of Infection. Attention has already been drawn by one of us, both in a research undertaken in conjunction with Lamb* and in a research undertaken in conjunction with Douglas,f to the fact that we have in the actual focus of * ‘Lancet,’ December 23, 1899. + ‘Roy. Soc. Proc.,’ vol. 74, p. 157. 206 Dr. A. E. Wright and Staff-Surgeon S. T. Reid. [Oct. 21, infection a lowered “bacteriotropic pressure” which accounts for the cultivation of the pathogenetic microbe in the interior of an organism which has at disposal in the circulating blood a considerable reserve of anti- bacterial substances. We propose here in conclusion to furnish further illustration of the general law as enunciated above, culling our examples not alone from the observations we have made in connection with tubercular infection, but also from observations made in connection with other bacterial infections. Observation 1.—Case of abscess in the neighbourhood of the appendix, Blood from the patient’s finger and pus obtained from the abscess at the operation were examined, with a view to determining the nature of the infection. Phagocytic counts. With a suspension of | With a suspension of tubercle bacilli. staphylococci. PSEISUER Socnoaccasssa0oGnod8ace. 2°3 45 Fluid obtained from the 0-1 1-9 pus by centrifugalisation The fact that the tuberculo-opsonie power of the patient’s blood was here 23 times as great as that of the fluid obtained from the pus was taken as evidence that tuberculo-opsonic substances had been used up in the pus and that the patient was suffering from a tubercular infection. It was inferred on similar grounds that he was also infected by staphylococcus. Observation 2.—Case of osteo-myelitis of the femur. Blood from the patient’s finger and pus obtained from the abscess at the operation were examined, with a view to determining the nature of the infection. Tuberculo-opsonic Staphylo-opsonic index. index. S{zPAUEIL. “Ggaqaqneqoo0090900000000 1:0 2°5 Fluid obtained from the 1-1 0-9 pus by centrifugalisation The fact that the opsonic index of the patient’s circulating blood was here normal to tubercle, while it was two and a-half times greater than normal 1905.] On Tubercular Infection in a Patient's Blood, etc. 207 with respect to the staphylococcus, was taken as evidence that the patient was not infected with tubercle, and that he was infected by staphylococcus, and had responded to that infection by a production of immunising substances. The fact that the tuberculo-opsonic index of the fluids obtained from the pus was the same as that of the blood, while the staphylo-opsonic power was only two-fifths of that of the circulating blood, was taken as of con- firmatory evidence of the conclusion already arrived at. The fact that a copious culture of staphylococcus aureus was obtained from the pus, planted out with aseptic precautions at the operation, further confirmed the diagnosis. Observation 3.—Case of psoas abscess. Blood from the patient’s finger and pus from the abscess were examined. Phagocytic counts. With a suspension of | With a suspension of tubercle bacilli. staphylococci. SEVIER, socgrasdscousondcupenede 2-4, 5°0 Fluid obtained from the 1°23 1:2 pus by centrifugalisation The fact that the fluid obtained from the pus was impoverished in both tuberculo- and staphylo-opsonic substances as compared with the blood was taken as evidence of a combined infection by tubercle bacilli and staphylo- cocci. This inference was confirmed by the fact that the opsonic power of the blood with respect to both the micro-organisms here in question was undergoing perpetual fluctuations.* The inference so far as it related to the staphylococcus was further confirmed by the fact that cultures of the micro- organism were obtained from the pus. Observation 4.—Case of ascites with grave constitutional disturbance in a man of 30. Blood from the finger and ascitic fluid were examined on two occasions. First Occasion. Tuberculo-opsonic index. DOWUMG ce Feces ove debited 1:05 ANSOMBOTINCL asec ancpese 0:99 We reported upon this that the patient was not suffering from tubercular peritonitis. * For the variations registered in connection with the tuberculo-opsonic power, vide supra, p. 196 of this paper, where Example 3 refers to the patient here in question. 208 Dr. A. E. Wright and Staff-Surgeon 8. T. Reid. [Oct. 21, The clinical symptoms, the age of the patient, and the appearances as seen at the operation appearing in contradiction with this verdict, and the ascites having reappeared, a second operation was performed, and a further sample of ascitic fluid was obtained for examination. At the same time the clinical appearances were again observed, with the result that there was now some wavering as to whether the original diagnosis could be upheld. The result of the phagocytic examination of the ascitic fluid, and of a sample of blood from the fingers were now as under :— Tuberculo-opsonic index. Seralt Giescecsaeme sveees 1 In view of this result the verdict previously given was sustained. A post-mortem examination, which followed in the course of a few weeks, again threw doubt on the verdict, the naked-eye appearances being entirely consistent with the theory of miliary tuberculosis affecting the peritoneum and serous covering of the intestines. Microscopic examination of the sections made through the miliary nodules revealed, however, a typical picture of miliary carcinoma. No primary carcinomatous focus had been discovered, though it was sought for on post-mortem examination. Observation 5.—Case of pleural effusion. Blood from the finger and fluid obtained by paracentesis of chest were examined :— Tuberculo-opsonic index. NOPE psp -eeeeapeenesene 0:92 Pleural (fluid) -s325-6-.2: 10 This was taken as evidence of the absence of tubercular infection. Observation 6.—Case diagnosed as peritoneal tubercular peritonitis com- plicated with pleurisy. Blood from the finger was examined on two occasions. On the second occasion, which was 48 hours after the first examination, peritoneal and pleural fluid were also examined. The results obtained by the phagocytic examination undertaken on this second occasion were as follows :— Tuberculo-opsonic index. Seruin jie yeasts 0:7 Peritoneal fluid ......... 0:28 Rleuralstiurdeseee eee il The results of the comparison of the peritoneal fluid with the serum obtained from the blood withdrawn from the finger were taken as evidence of tubercular infection of the peritoneum. Confirmatory evidence of tubercular infection was furnished further by the low tuberculo-opsonic 1905.| On Tubercular Infection in a Patient's Blood, etc. 209 power of the blood, and by the observed fluctuation in this index. When it was measured two days previously, this index had worked out as 1°4. The fact that the opsonic power of the pleural fluid worked out as higher than the opsonic power of the serum was taken as evidence that the pleural effusion had occurred at a period when the opsonic power of the blood was 1 or above 1. The diagnosis of tubercular infection of the peritoneum and pleura (and underlying lung) was confirmed at the post-mortem examination. Observation '7.—Case of long-continued suppuration of the antrum associated with the presence in the pus of the pneumococcus and the Bacillus fusiformis and Spirillum bucee of Vincent. The patient had been treated by therapeutic inoculations of a pneumococcus vaccine. The patient’s serum and the antral pus were examined with a view to determining whether the pneumococcus played any active part in connection with the continuance of the suppuration :-— Pneumo-opsonic index, SOMITE ete ida. a MAIR we'side nM eM Maes bot debininc tceei we 4:3 Fluid obtained from pus by centrifugalisation ...... 0°3 The results were taken as evidence (a) that the pneumococcus played an active réle in connection with the suppuration, and (0) that the protective substances which had been generated in the blood under the influence of inoculation did not come satisfactorily into application upon the micro-organisms in the antrum. Observation 8.—Case of whitlow associated with the formation of a blister under the nail. Serum derived from blood from a sound finger and blister fluid were examined. Staphylo-opsonic index. SEIT Bbaboecodeanse see A 0:8 IBiStera shld weemeeese eee 0:3 The blister fluid yielded a pure culture of staphylococcus. Observation 9.—Rabbit in the early stages of anthrax infection —Blood obtained from the ear and lymph from the seat of inoculation were examined. Anthraco-opsonic index.* SOLU ya chest eat ans duetue Tey MyM soos. showeaesecseper 0°62 * Tested with a suspension of anthrax spores and compared with the serum of a normal rabbit tested in the same mannet. It may be noted that all the difficulties and inaccuracies which are associated with the employment of ordinary anthrax cultures in phagocytic experiments can be satisfactorily evaded by the employment of suspensions of anthrax spores. These, when stained with carbol fuchsin and decolourised by 0°25 per cent. sulphuric acid, represent absolutely ideal elements for enumeration. 210 On Tubercular Infection in a Patient’s Blood, etc. APPENDIX. A further Series of Observations showing that Phagocytosis is obtained with the Heated Serum of Patients who are the subjects of a Systeme as distinguished from a strictly Localised Tubercular Infection, or who, being the subjects of a strictly Localised Tubercular Infection, have been subjected to Inoculations with Tubercle Vaccine. The serum was in each case heated to 60° C. for 10 minutes. Table supplementary to Table I1—Showing that the Normal Serum, after it has been exposed to a Temperature of 60° C. for 10 minutes, no longer incites Phagocytosis. Unheated serum. Heated serum. : Opsonie ae index Phagocytic (determined Phagocytic Gee ee . count. by compari- count. y basse Pali - (Number of son of (Number of phaede sie number i F bacteria hagocytic bacteria of obser- Natnre of infection, ingested divided 5 cuit ingested divided are t vation. by number of with that by number of be : 4 leucocytes obtained leucocytes PA a examined.) with pooled examined.) we he q blood of bl i ne healthy men). ea 2 normal men). 1 Fibroid phthisis, tu- | (100/30) = 3:3 1-0 (142/37) = 4:0 1:2 bercle bacilli in sputum 2 | Early phthisis, tu-| (132/80) =4'4| 1°83 (122/47) = 2°6 0°77 berele bacilli in sputum 3 Acute phthisis, tu- | (180/80) = 4:3 1°3 (96/40) = 2°4 0°74, bercle bacilli in sputum 4 | Acute phthisis ......... (127/40) = 3-2 1-0 (45/34) =1°3| 0-4 5 | Fibroid phthisis (?) ...| (182/30) = 6 0 1°8 (51/48) = 1-2 0:3 6 Phthisis, tubercle | (117/30) = 3°9 1:1 (65/30) = 2-2 0-62 bacilli in sputum 7 Mitral stenosis ......... (106/30) = 3°5 1°0 (19/31) = 0°6 0:17 8 Early phthisis ......... (161/30) = 5°4 1°6 (54/27) = 2°0 0°6 Q | Phthisis .......c..cc0000 (257/40) = 6°4| 1:3 51/40) =1°3| 0:27 10 Lupus under treat- | (131/36) = 3:3 16 (74/40) = 1°8 0°'8 ment by inoculation of tubercle vaccine 11 Lupus under treat-| (73/380) = 2°4 1:2 (81/80) = 1-0 0°5 ment by inoculation of tubercle vaccine 12 Tubercular ulcer of | (63/30) = 2:1 1:2 (60/30) = 2:0 11 leg under treatment by inoculation of tubercle vaccine The first eight of the observations here in question were made upon bloods collected for us in the Victoria Park Hospital by our fellow worker, Dr. G. W. Ross. The clinical diagnosis which had been arrived at was not made known to us till afterwards, when the particulars set forth in Column 2 were filled in by Dr. Ross. On Spontaneous and other Phagocytosis. 211 In contrast with the observations incorporated in Table II in the body of the paper these observations were conducted in phagocytic mixtures containing 0°85 per cent. instead of 1:1 per cent. of NaCl. It is shown in the next following communication that spontaneous phagocytosis is absolutely abolished only in the case when the salt content of the phagocytic mixture exceeds 1 per cent. The source of fallacy to which attention is here called falls, no doubt, for all practical purposes, entirely out of account. On Spontaneous Phagocytosis, and on the Phagocytosis which is Obtained with the Heated Serum of Patients who have Responded to Tubercular Infection, or, as the case may be, to the Inoculation of a Tubercle Vaccine. By A. E. Wricut, M.D., sometime Professor of Pathology, Army Medical School, Netley, Pathologist to St. Mary’s Hospital, London, W., and Staff-Surgeon 8. T, Rem, R.N. (From the Pathological Laboratory of St. Mary’s Hospital, London, W.). (Communicated by Sir John Burdon-Sanderson, Bart., F.R.S. Received October 21,—Read November 23, 1905.) It has been indicated in the foregoing paper than an incitor element* is to be found in the blood of those who have made an immunising response to tubercular infection, or, as the case may be, to an inoculation of a tubercle vaccine. This fact does not stand by itself. Recital of Previous Observations on the same Subject. The observations of Metehnikoff, following in sequence upon the classical researches of R. Pfeiffer on the intraperitoneal destruction of bacteria by the aid of immune sera, first drew attention to the fact that very active phagocy- tosis comes under observation when bacterial cultures, or as the case may be Spermatozoa, are introduced into the peritoneal cavity of normal animals in association with heated serum derived from immunised animals. * The term “incitor-element” (Latin—zncito: I hasten, I urge forward, I bring into rapid movement) is here employed to denote the element in the heated serum which promotes phagocytosis. By employing this term, pending the elucidation of the nature and mode of action of the element in question, we secure the advantage of leaving these issues unprejudged. 212 Dr. A. E. Wright and Staff-Surgeon 8. T. Reid. [Oct. 21, Savtschenko* obtained in experiments conducted in vitro with the heated sera of animals which had been subjected to injections of red blood corpuscles, phagocytosis of these formed elements. Neufeld and Rempau,t working with heated sera derived from animals hel had been immunised against streptococcus and pneumococcus, and conducting their experiments in vitro, have described these immune sera as possessing a power of inciting phagocytosis. This power was, be it remarked, not numerically measured. Leishman,t employing the numerical method for the measurement of phagocytosis which was devised by him with the modifications introduced by one of us in conjunction with Douglas, ascertained that the sera derived from Malta fever convalescents, or as the case may be from men who had undergone anti-typhoid inoculation, retain, after heating, elements which promote phagocytosis. Dean, working with the same methods, without however conforming to the easily realised conditions§ which are essential to the accuracy of the enumeration, has described incitor elements in the heated serum derived from animals which had been immunised against staphylococcus. Lastly, Douglas, employing again the same methods, has obtained evidence of the presence of an incitor element in the heated serum derived from himself and others after inoculation with a sterilised culture of the plague bacillus. Views of the Observers above mentioned on the Nature of the Incitor Element contained in the Heated Serum. Influenced by the theoretical conception that the increased resistance to bacterial invasion which is obtained by bacterial inoculation is in every case referable to a modification of the phagocytes,|| Metchnikoff originally spoke of the incitor element as a stemulin. * © Annales de l'Institut Pasteur,’ 1902. + Neufeld and Rimpau’s paper was published in the ‘ Deutsche Med. Wochenschrift’ in September, 1904, 12 months after the first description of the opsonins in these ‘ Proceedings.’ t ‘Path. Soc. Trans.,’ 1905, vol. 56. § “I should not feel disposed,” remarks this author (‘ Roy. Soc. Proc.,’ Series B, vol. 76, p. 511), “to place quite the same reliance as Wright and Douglas on the numerical accuracy of the results which can be derived from their method. Where the leucocytes are very full, z.¢., where the counts are high—it is impossible to differentiate results by the method of enumeration.” In spite of the perfectly self-evident experimental limita- tion of our method, which Dean here recognises, this worker employs in practically all his published experiments bacterial suspensions which give him an average phagocytic count often of 50 and more bacteria in the leucocyte. Such a count is altogether incompatible with accurate quantitative work. || The correctness of the view that artificial immunity depends upon a modification of 1905. | On Spontaneous and other Phagocytosis. 213 This appellation may, we think, be characterised as unfortunate, jirst, because the mode of action of the incitor element was prejudged ; secondly, because the appellation suggests (in contravention to everything which has come to light with respect to immunisation) that there are elaborated in the animal organism in response to inoculations, not vaccinotropic elements (elements which have a chemical affinity for the vaccine) but lewcocytropic elements (substances which have a chemical action on leucocytes). At a later date the terms “sensitiser ” and “fixing substance” (Ja substance sensibilitrice and le fixateur) were applied by Metchnikoff to the incitor element. This nomenclature is, it seems to us, almost equally infelicitous— infelicitous because it imposes upon the mind the following ideas :—(q) that the phenomena of phagocytosis are analogous to those of hemolysis ; (>) that the incitor substance, like the “ amboceptor” of Ehrlich, exerts its specific effect only in the case where it is reinforced by a complement; and (c) that the mechanical movements of the phagocyte in the ingestion of particulate matter are analogous to the chemical action of the complement in the case where red blood corpuscles are dissolved by a hemolytic serum. With the exception of Leishman,* who, with a view to conforming to the original nomenclature of Metchnikoff, and also because his own experi- ments incline him to adopt the same point of view, speaks of the incitor substances as stimulins, all the other observerst take the view that the the leucocytes was first inquired into by Denys and Leclef (‘ La Cellule,’ 1895, vol. 11), in connection with their experiments conducted on rabbits with streptococcus. The doubt with regard to the correctness of Metchnikoff’s view which found expression in the paper of these authors was further justified by the experiments of Mennes (‘ Zeitsch. f- Hygiene, 1897, vol. 25), conducted with the blood of animals immunised against the pneumococcus. Finally, the incorrectness of the view that immunisation depends on a modification of the leucocytes was for the first time unambiguously established by one of us working in conjunction with Douglas (‘ Roy. Soc. Proc.,’ vol. 72, p. 369, and vol. 73 p.129). Our results were afterwards confirmed by Bulloch (‘ Roy. Soc. Proc.,’ vol. 75). * Loe. cit. and ‘ Journ. of Hygiene,’ 1895. + It may be remarked in this connection that Neufeld and Rimpau, while satisfied that the incitor substances in the serum exert an opsonic action on the bacteria, suggest that the term opsonins should be here rejected and that the substances here in question should be called bacteriotropins. Pending the discussion of the questions of the mode of action of the incitor elements in the heated serum, and of their identity or non- identity with the opsonins found in normal blood, it will suffice here to remark with respect to the proposed nomenclature of Neufeld the following :— (a) The term bacteriotropins (since it connotes nothing more than the property of entering into chemical combination with bacteria) is more appropriate as a generic term for the whole class of substances which combine chemically with bacteria, than as a specific designation for the substances which prepare the bacteria for phagocytosis. (6) All considerations of the comparative merits of Neufeld’s terminology and my terminology apart—there must, I apprehend, remain to me as the author of the term VOL. LXXVII.—B. Q 214 Dr. A. E. Wright and Staff-Surgeon S. T. Reid. [Oct. 21, incitor element in the immune serum exerts an opsonic action upon the bacteria, preparing them for phagocytosis. Sourees of Fallacy which must be Eliminated before the Question as to the Nature of the Incitor Element in the Heated Serum can be Properly Investigated. Before an inquiry into the nature of the incitor constituent of heated “immune serum” can be properly taken in hand, the sources of fallacy which are incident to such an inquiry must be realised. A first source of fallacy is associated with the occurrence of spontaneous phagocytosis. 3000/1. Fic. 4.—Diagram of a somatic amphiaster, in which longitudinally split V-shaped chromosomes, with limbs of unequal length, are apparently arranged parallel to the spindle axis. Adjacent chromosomes, with their longer limbs on opposite sides of the equator, if regarded as together forming one chromosome, would convert such a mitosis into a heterotype with half the somatic number of chromosomes arranged longitudinally on the spindle, e.g., figs. 3, 5, 6, 7, 8, 9, and 10. Fies. 5 and 6.—Apparent heterotypical mitosis. Fig. 5, replica of fig. 4, Royal Society paper, and of fig. 26 in First Scientific Report, 1904. Transplanted carcinoma of * ©Centralb. f. Path. u. path. Anat.,’ vol. 18, 1902. + Loe. cit. 232 On the Occurrence of Heterotypical Mitoses in Cancer. mouse. Chromosomes arranged longitudinally on the spindle. The mitosis is contained in two consecutive sections. 3000/1. Fies. 7 and 8.—Same sections as figs. 5 and 6. Result of analysis after restaining. Fic. Fie. Longitudinally split chromosomes with unequal limbs projecting above and helow the equatorial plane. X 3000/1 of diagram, fig. 4. 9.—Apparent heterotypical mitosis. Transplanted carcinoma of mouse. Loop and figure-of-8 chromosomes arranged longitudinally on the spindle. > 3000/1. 10.—Analysis of the same preparation as fig. 9, showing the slight differences in interpretation sufficient to make this mitosis conform to the somatic type. The loop chromosome in the middle of the equatorial plate consists of two distinct V-shaped .chromosomes with unequal limbs projecting above and below the equator. The Fic. Fic. PIG. Fic. attraction fibres are attached to the apices, and not to the ends of the long limbs as would be the case in a true heterotype. X 3000/1. 3. 11.—Diagram of a somatic metaphase in which the limbs of the chromosomes are of unequal length. The longer limbs still cohere after separation of the apices and shorter limbs. The barrel-shaped figure thus produced resembles a heterotype, especially when the compressed form of the cytoplasm crowds the chromosomes together. 12.—Shows the detailed analysis of the mitosis at the upper part of fig. 20, Plate 7, Second Scientific Renort, 1905. It illustrates the mode of separation of daughter chromosomes with unequal limbs, as represented diagrammatically in fig. 11. Transplanted carcinoma of mouse. X 3000/1. 13.—Microphotograph (untouched) of “monaster” mitosis from squamous-celled carcinoma of the tongue (man). Shows ring and U-shaped chromosomes. 1000/1. 14.—Analysis of same section as fig. 13. Partial separation of the daughter chromosomes accounts for the presence of ring and U-shaped chromosomes. No centrosomes or achromatic figure visible. X 3000/1. 15.—Remainder of same cell in next section. Shows large number of chromosomes of ring and U-shape, along with others in which the widely separated daughter-rods are parallel to each other. x 3000/1. Bashford and Murray. Roy. Soc. Proc., B. vol. 77, Plate 5. 7 Bashford and Murray. Roy. Soc. Proc., B. vol. 77, Plae 6. 233 Pathological Report on the Histology of Sleeping Sickness and Trypanosoniasis, with a Comparison of the Changes Found am Animals Infected with T. Gambiense and other Trypano- somata. By Anton BreEINL, M.U.Dr. (Prague), J. W. Garret International - Fellow, Liverpool School of Tropical Medicine. University, Liverpool. (Communicated by Professor R. Boyce, F.R.S. Received April 8,—Read May 11, 1905.) Three cases of Sleeping Sickness and one case of Trypanosomiasis dying in Liverpool have been histologically examined. The central nervous system of the Sleeping Sickness cases showed the changes described by different observers, Mott, Low, the Portuguese Commission and others. One case exhibited an intra-pial hemorrhage of the spinal cord, extending from the sixth cervical segment to the third thoracic segment, about 7 mm. broad. In another case there occurred four larger hemorrhages, besides numerous smaller ones, in the grey substance, chiefly affecting the posterior cornua and the thoracic part of the cord. Microscopically the brain and spinal cord showed small celled infiltration around the vessels, consisting for the most part of lymphocytes, some plasma cells and phagocytes, between which were a varying number of red cells in different stages of disintegration. The intima of the vessels showed a proliferation of the endothelial cells. Red and white blood corpuscles were often seen in the vessel walls. Here and there the blood vessels were filled with white blood corpuscles resembling a thrombosis. It is most striking that the small celled infiltration is much more marked in the grey substance of the nervous centres, especially in the large grey ganglia, than in the peripheral parts. Very numerous capillary hemorrhages of different sizes were present in these situations. Infiltration around the vessels of the membranes and in the tissues of the pia and arachnoidea was observed. Around the infiltrated vessels degeneration of the fibres and an excess of glia cells were seen, sometimes exhibiting the picture of red softening. The ganglia cells showed an _ irregularly distributed degeneration, central and peripheral chromatolysis and also partial pyknosis. Signs of inflammation and small celled infiltration im the endo- and peri- 234 Dr. A. Breinl. On the Histology of [Apr. 8, neurium of the peripheral nerves were seen. In two cases of longer duration and with more pronounced symptoms of the disease, more definite changes around the vessels of the brain and spinal cord were seen than in the third case of shorter duration with less marked symptoms. In this case the peri- vascular changes in the brain were small and were still less so in the spinal cord. In all groups of the lymph glands numerous ones were found showing the typical appearance of hemo-lymph glands with a pronounced hyperplasia of the connéctive tissue, a widening of the follicles and the formation of a system of sinuses containing red blood cells and large phagocytes in a fine threadwork of connective tissue. Others showed a transition between the heemo-lymph glands and normal glands, one part appearing normal, the other presenting a typical sinus formation with numerous red blood cells and phagocytes. Nearly all the glands contained between the lymph cells a number of blood corpuscles, many in all stages of degeneration. The spleen was greatly congested and contained a few necrotic areas, scattered through the organ was a little blood pigment giving the iron reaction. The bone marrow exhibited the typical picture of red marrow with gelatinous degeneration. The liver and kidneys showed hemorrhages between the parenchyma cells, which latter appeared to be undergoing degeneration. In all three cases a few large bacilli and cocci were seen which did not stain by Gram’s method, these I consider to be due to post-mortem contamination. The bacteriological cultivation, anaérobic and aérobic of the cerebro-spinal fluid and the blood of two cases, did not give any growth, and moreover, animals infected with large quantities of cerebro-spinal fluid or:blood did not show any other symptoms than those caused through the presence of trypanosomes in the blood. In only one case, dying with a fair number of trypanosomes in the blood, could I find occasionally a parasite in the congested vessels of the organs. The one case of trypanosomiasis, which died from an intercurrent pneumonia, did not show any other changes in the central nervous system than the very large peri-vascular spaces, partially filled with transudate, and sometimes containing a few white blood corpuscles. The ganglion cells showed the changes corresponding to the hyperthermia. The lymph glands were very hemorrhagic, some showed the typical appearance of heemo- lymph glands. The brains, spinal cords, ana organs of numerous animals infected with Trypanosoma Gambiense, monkeys, rabbits, guinea-pigs, dogs, rats, and mice were examined. One of the monkeys showed a typical hemorrhagic cicatrix in the left lobus centralis of the brain; other monkeys and a chimpanzee 1905. | Sleeping Sickness and Trypanosomiasis, ete. 235 showed a high congestion of the vessels of the brain and spinal cord, with hemorrhages, around the vessel walls, containing lymphocytes, a few leucocytes, and phagocytes. The intima showed large proliferated endo- thelial cells, the vessels often contained very many leucocytes. Numerous hemorrhages in the grey substance of the spinal cord were frequently seen. Some of the dogs, rabbits, and guinea-pigs showed the changes in the spinal cord, and to a less extent in the brain. The ganglion cells exhibited similar alteration as in the human cases. In some of the animals no changes around the vessels and very little alteration of the ganglion cells and fibres were noted. Many of the lymph glands presented the picture of hemo-lymph glands with a few pigment granules; sometimes an irregular patchy appearance was seen, the centre consisting of a light stained area with numerous red cells and phagocytes, the periphery of normal lymph tissue with a small number of follicles. The spleen showed congestion in the more acute cases, with irregular hyperplasia of the malpighian bodies, in the older cases hyperplasia of the connective tissue. For comparison the brains, cords, and organs of animals infected with 7. dimorphum (Gambian horse disease) were examined. In a few cases the same hemorrhages as described above and localised in the grey substance of the nervous centres were seen. The lymph glands showed the peculiar appearance ; as noted above the light spaces were completely filled with blood pigment. The spleen showed hardly any pigment. Trypanosomes were tound mostly clumped together in the vessels of the different organs of all animals dying with numerous parasites in the peripheral blood. Conclusions. (1) In the cases of Sleeping Sickness there is a pronounced congestion of the blood vessels of the central nervous system ‘together with a small celled infiltration around the vessels of the brain and spinal cord, especially in the grey substance. (2) Chromatolysis and pyknosis of the ganglion cells of brain and spinal cord. (3) Inflammation of the leptomeninges of the brain and spinal cord. (4) Neuritis of the peripheral nerves. (5) The more chronic the case and the more pronounced the symptoms the greater the changes in the brain and cord. (6) The majority of the lymph glands exhibit the picture of heemo-lymph glands. (7) Small necroses of the spleen and signs of degeneration of the bone marrow. 236 © Mistology.of Sleeping Sickness and Trypanosomaasis, ete. (8) The brain of a case of Trypanosomiasis did not show small celled infiltration. (9) Animals infected with Trypanosoma Gambiense show sometimes changes in the nervous system, localised in the grey matter, hemorrhages, lymphocytes, and a few leucocytes in the peri-vascular space: hemo-lymph glands in large nuinbers, and sometimes necrosis of the spleen and degenera- tion of the bone marrow. (10) Animals infected with 7rypanosoma dimorphum exhibit similar changes in the nervous system and organs. A far greater deposit of pigment in the lymph glands and in older cases in the spleen is present. 1905. | Sleeping Sickness and Trypanosomiasis, etc. 235 showed a high congestion of the vessels of the brain and spinal cord, with hemorrhages, around the vessel walls, containing lymphocytes, a few leucocytes, and phagocytes. The intima showed large proliferated endo- thelial cells, the vessels often contained very many leucocytes. Numerous heemorrhages in the grey substance of the spinal cord were frequently seen. Some of the dogs, rabbits, and guinea-pigs showed the changes in the spinal cord, and to a less extent in the brain. The ganglion cells exhibited similar alteration as in the human cases. In some of the animals no changes around the vessels and very little alteration of the ganglion cells and fibres were noted. Many of the lymph glands presented the picture of hemo-lymph glands with a few pigment granules; sometimes an irregular patchy appearance was seen, the centre consisting of a light stained area with numerous red cells and phagocytes, the periphery of normal lymph tissue with a small number of follicles. The spleen showed congestion in the more acute cases, with irregular hyperplasia of the malpighian bodies, in the older cases hyperplasia of the connective tissue. For comparison the brains, cords, and organs of animals infected with 7. dimorphum (Gambian horse disease) were examined. In a few cases the same hemorrhages as described above and localised in the grey substance of the nervous centres were seen. The lymph glands showed the peculiar appearance ; as noted above the light spaces were completely filled with blood pigment. The spleen showed hardly any pigment. Trypanosomes were found mostly clumped together in the vessels of the different organs of all animals dying with numerous parasites in the peripheral blood. Conclusions. (1) In the cases of Sleeping Sickness there is a pronounced congestion of the blood vessels of the central nervous system together with a small celled infiltration around the vessels of the brain and spinal cord, especially in the grey substance. (2) Chromatolysis and pyknosis of the ganglion cells of brain and spinal cord. (3) Inflammation of the leptomeninges of the brain and spinal cord. (4) Neuritis of the peripheral nerves. (5) The more chronic the case and the more pronounced the symptoms the greater the changes in the brain and cord. (6) The majority of the lymph glands exhibit the picture of hemo-lymph glands. (7) Small necroses of the spleen and signs of degeneration of the bone marrow. VOL, LXXVII.—B. 5 236 Messrs. Bateson, Saunders, and Punnett. | Dee. 1, (8) The brain of a case of Trypanosomiasis did not show small celled infiltration. (9) Animals infected with Trypanosoma Gambiense show sometimes changes in the nervous system, localised in the grey matter, hemorrhages, lymphocytes, and a few leucocytes in the peri-vascular space: hemo-lymph glands in large numbers, and sometimes necrosis of the spleen and degenera- tion of the bone marrow. (10) Animals infected with 77rypanosoma dimorphum exhibit similar changes in the nervous system and organs. A far greater deposit of pigment in the lymph glands and in older cases in the spleen is present. Further Experiments on Inheritance in Sweet Peas and Stocks : Prelimanary Account. By W. Bateson, F.R.S., E. R. SAUNDERS, and R. C. PUNNETT. (Received December 1,—Read December 7, 1905.) Later results have provided expressions which include many of the peculiar phenomena of inheritance already witnessed in sweet peas and stocks. In sweet peas we have shown that purple may occur, as a “reversion,” from the cross between two whites, one having long pollen grains, the other round. Similarly in stocks, white glabrous x cream glabrous gives “yeversionary”” F, purple hoary. (In both cases the parents are whites, a.¢., tree from sap-colour, for cream is due to yellow plastids, recessive to colourless plastids.) The appearance of coloured flowers is due to the simultaneous presence in the zygote of two factors, belonging to distinct allelomorphic pairs, which may be spoken of as C, ¢, and R, 7, the large letter denoting presence, the small letter the absence of the particular factor. Hoariness of stocks is similarly due to the-coexistence of two other factors, and the presence of either of these factors is also allelomorphic to its absence. These two pairs are spoken of as H, h, and K,/. But, though H and K may both be present, no hoariness is produced unless C and R, the colour-factors, are also both present. For the actual development of hoariness four factors are thus required. The existence of white-flowered hoary plants creates a difficulty; but white incana is evidently a coloured 1905. | Inheritance in Sweet Peas and Stocks. 237 form in reality, for its flowers tinge on fading, and its embryo has the deep- green colour characteristic of purple varieties. Apart from breeding-tests, however, white hoary Bromptons show no visible indication of colour, and as yet they constitute a marked exception. White glabrous and cream glabrous types contain both H and K, the two elements of hoariness. One of them contains C and the other contains R. All sap-coloured types studied contain one only of the two factors H, K. Consequently, we find the following result, which formerly seemed paradoxical :— we 1. Cream glabrous x Red or purple’ Red or purple hoary. glabrous 2, White glabrous x Ditto Purple hoary. 3. Cream glabrous x White glabrous Ditto. 4. Any red or purple x Any red or purple Red or purple glabrous. glabrous. glabrous. The truth of this account appears from the fact that in F, from cream glabrous x white glabrous all the coloured are hoary and all the whites are glabrous. Again, purple (hoary) incana x cream glabrous gives in F, all the hoary plants colowred, and all the glabrous plants white; while “white” (hoary) incana x sap-coloured types gives in F, coloured hoary, coloured glabrous, and in addition tinging “ whites” in both classes. When a character is produced by the meeting of factors belonging to two distinct allelomorphic pairs, the F2 ratio will be 9:7 (a2, 3+3+1), and consequently, when in sweet peas and stocks a coloured F, is produced from two non-sap-coloured types, the F2 ratic is 9 coloured : 7 white; but there are 4 gametically-distinct types among the coloured and 5 among the whites. Most of these have been now recognised experimentally. When F; is purple the coloured class consists of purples and reds. In both sweet peas and stocks the ratio is 27 purple, 9 red, 28 white, composed thus :-— 4s O2eOsOs Se Seoes oil The purples are due to the presence of a “blue” factor B, allelomorphic to , its absence. Unless C and R are both present, B cannot be perceived without breeding tests. The three pairs, C,c, R, 7, B, b, by entering into all possible combinations according to the simple Mendelian system, give the results observed. This scheme takes no account of the sub-classes which sometimes occur Ss 2 238 Inheritance in Sweet Peas and Stocks. in both purples and reds. Several of these are merely superposed on the primary classes, while others are more complex and require further analysis. The distribution of the colours shows further complications when some coloured strains were introduced as original parents. “Reversion” is thus seen to be a simple and orderly phenomenon, due to the meeting of factors belonging to distinct though complementary allelo- morphic pairs, which at some moment in the phylogeny of the varieties have each lost their complement. Pollen-characters in Sweet Peas.—Gametic coupling of a novel kind exists in this case. The whole generation in Fy, consists of 3 long: 1 round. The whites taken alone also are 3 long: 1 round. But in the purples there is a great deficiency of rounds, while in the reds they are greatly in excess. This result indicates that there is a partial coupling of the long pollen- character with the factor B, and a corresponding coupling of round pollen with 06. This peculiarity only occurs in families which contain both purple and red members. The gametic output of F, in these cases is approximately 7AB+1A6+4+ 1aB+7ab, where A is long, and a round pollen. This arrangement gives a close approach to the observed figures :— Purple. Red. White. (Re ee ea Sa ee Sepa Long. Round. Long. Round. Long. Round. Observed ...... 1528 106 7, 381 1199 394 Calculated ...... 1448-5 122°7 122;7' 4015 12205) 407s 239 Report on the Psychology and Sociology of the Todas and other Indian Tribes. By W. H. R. Rivers, M.D., Fellow of St. John’s College, Cambridge. (An Abstract of Work carried out by the aid of the Gunning Fund of the Royal Society for the year 1901-1902. Communicated by the Secretaries of the Royal Society. Received October 18,—Read December 14, 1905.) Six months were spent in India, the greater part of the time being devoted to the investigation of the Todas of the Nilgiri Hills. The senses of these people were examined experimentally on the same lines as those followed by the Cambridge Expedition to Torres Straits.* The general result was to confirm the chief conclusion of this expedition that there are no great differences between the senses of savage and civilised races. In pure sense-acuity little difference was found, and the observations lend no support to the view that the sense-acuity of savage or harbarous races is superior to that of civilised man, the apparent superiority in some cases being due to the training of observation in special directions. In two senses only is there distinct evidence of difference between Todas and Englishmen in sensory endowment. The Todas are distinctly less sensitive to pain than the average educated Englishman, and they show the same kind of deficiency in the colour-sense which has been found in other races of low culture, especially in the Papuanft and the Egyptian peasant.t The Todas are distinctly less sensitive to blue than the average educated Englishman, though differing little in sensibility to red or yellow. This defect in the sensibility for blue is associated with the deficient nomen- clature for this colour which is almost universal in races of low culture; and the observations on the Todas strengthen the conclusion reached by previous work that physiological insensitiveness is one, though only one, of the factors upon which the defect in language depends. The most striking feature of Toda colour-vision, however, is the great frequency of colour-blindness. About five hundred individuals were tested, and over 12 per cent. of the males were found to suffer from typical red- green blindness, the proportion in European races being about 4 per cent. In most races of low culture colour-blindness is less frequent than in * ‘Reports of the Cambridge Anthropological Expedition to Torres Straits.’ Cambridge, vol. 2, Part I, 1901, and Part II, 1903. t Loe. cit., p. 48. t ‘Journ. Anthrop. Inst.,’ 1901, vol. 31, p. 229. 240 Psychology and Sociology of the Todas, ete. Europe, but the Todas show the highest recorded frequency of this condition in any race. By means of the genealogies preserved by the Todas the relationship between the colour-blind people could be traced, and a body of material was obtained which illustrates the mode of hereditary transmis- sion of the defect. . The Todas were found to be subject to various geometrical-optical illusions ; and quantitative observations were made on the illusion of compared vertical and horizontal lines and on the Miiller-Lyer illusion. The Todas are subject to the former in a greater degree than English observers, and to the latter in a smaller degree. The two illusions differ in nature: the former is probably largely physiological in origin, and is neutralised by the experience of civilised life, while the latter is more strictly psycho- logical in character; and the different reaction of the Todas to the two illusions is in accordance with this difference in their nature. In every measurement the degree in which the individuals of each race differed from one another was studied; and a mass of material was collected for the study of variability in the reaction to psychological tests, and for the analysis of the complex conditions upon which the coefficients of variation depend. On comparing the observations of Todas, Papuans, and Englishmen, all tested by the same methods, and chiefly by the same experimenter, it is found that there is some evidence of a correlation between the degree of general intellectual development and certain simple mental properties or activities which can be tested by experimental methods. In general intellectual development the Todas occupy an intermediate position between Papuans and Englishmen, and a similar intermediate position is occupied by them in connection with many of the tests.* The social and religious institutions were also studied. The sociology was investigated largely by means of the genealogical method,f and the system of kinship, the complex marriage regulations and the laws of inheritance and property were worked out in detail. The Todas were found to possess a highly elaborate religious ceremonial of which only brief sketches had previously been published, while many ceremonies had wholly escaped observation. In consequence, much time was devoted to the detailed investigation of this ceremonial and of the other features of the Toda religion. Evidence is given that this religion is one which has undergone degenerative changes, and some evidence is advanced * A full account of the senses of the Todas will be published shortly in the ‘ British Journal of Psychology, vol. 1, Part IV. + ‘Journ. Anthrop. Inst.,’ 1900, vol. 30, p. 74. Nitrification with reference to the Purification of Sewage. 241 in favour of a view that the Todas are a people who have once had a culture higher than that they now possess. When the customs and institutions of the Todas are compared with those of other parts of India, it is found that there is most resemblance with the people of Malabar; and the view is advanced that the Todas migrated to the Nilgiri Hills from Malabar, and are possibly allied in race to the two chief castes at present existing in that district, the Nairs and Nambutiris. In addition to the work on the Todas, observations were also made on members of other tribes. The vision of the Sholagas and Uralis, two wild jungle tribes, was investigated* from several points of view; and observa- tions, chiefly on colour-blindness, were made on members of other castes or tribes. A Study of the Process of Nitrification with reference to the Purification of Sewage. By Harrrerre Cuicx, D.Sc. (Communicated by Professor H. Marshall Ward, F.R.S. Received April 1,—Read May 11, 1905.) Introduction —That nitrification is a biological process was first established with certainty, after long controversy, in 1888, by the decisive experiments of Platht and Landolt, who in this matter confirmed the previous researches of Schlésing and Muntz,§ Warington|| and Soyka.{ The discovery of the active living agents followed soon after, when Winogradsky** (1890 to 1892) isolated the two sets of organisms which, as he showed, co-operate to produce natural nitrification. These were (1) the nitrite-producer, b. nitrosonvonas, which oxidises ammonia to the nitrite stage only, and (2) the nitrite-producer, &. nitrobacter, which carries on the * ‘Bull. Madras Government Museum,’ 1903, vol. 5, p. 3. + Plath, ‘Landw. Jahrbiicher, v. H. Thiel, vol. 16, hft. 6, and ‘Centralbl. f. Agri- kulturchem. v. Biedermann,’ vol. 17, 1888. t Landolt, ‘Deutsch. Landw. Presse, vol. 15, and ‘Centralbl. f. Agrikulturchem., vol. 17, 1888. § Schlésing and Muntz, ‘Comptes Rendus,’ vols. 84 and 85, 1877, and vol. 89, 1879. || Warington, ‘Journ. Chem. Soc., vol. 33, 1878, and ‘Landw. Versuchsst.,’ vol. 24, 1880. 4 Soyka, ‘ Zeitschr. f. Biologie, vol. 14, 1878. *= Winogradsky, ‘Ann. de l’Inst. Past., vol. 4, 1890, and vol. 5, 1891; also ‘ Archives des Sci. biol. de St. Petersb.,’ vol. 1, 1892. 242 Dr. H. Chick. The Process of Nitrification [Apr. 1, oxidation to nitrate but cannot act upon ammonia, being indeed inhibited mn its development by minute traces of that substance. Winogradsky, by him- self, and in conjunction with Omeliansky,* subjected these bacteria to a very exhaustive study. The most striking characteristic that they demonstrated was the marked repugnance of both forms to organic substances. Not only, in opposition to the rest of the plant world, do these organisms make no nutritive use of sugars, peptones, etc., but the presence of more than a trace of such organic substances was found to entirely inhibit their development, thus explaining the failure of all attempts to isolate these bacteria by using the ordinary nutrient culture-media:; Winogradsky, on the other hand, had succeeded in cultivating them by employing a silica-jelly-medium impregnated with inorganic salts, and a total imability to grow on organic nutrient media was afterwards put forward by him as a definite practical criterion of the purity of cultures of nitrifying bacteria. This criterion has been challenged by Burri and Stutzer,t by Stutzer and Hartleb,t and later by Fremlin.§ It has been shown by Winogradsky,| and also by Girtner, Frankel, and Kriigerf that the former workers were misled by an admixture of non- nitrifying organisms. The most recent work, that of Boulanger and Massol,** and of Wimmer tf confirms Winogradsky’s criterion. The special case of nitrification considered in this paper is that occurring during sewage purification, which aims at the complete oxidation and mineralisation of putrescible substances present. Nitrification is here of great importance, and the effluent of perfectly-treated sewage should contain all its nitrogen in the form of nitrates. Although land-treatment of sewage is theoretically the most economical, yet artificial processes, by which space can be saved, have often to be employed. Two processes concern us here, both involving the use of “filter-beds” of coke or other porous material, in which the sewage, usually after having been treated in a “septic tank” is oxidised by bacteria. (1) Contact Filters. —In these the filter-bed is first entirely filled up with the liquor and then after a time allowed to empty slowly, and finally to remain empty for a period. This cycle usually occupies about eight hours, and often may have to be repeated before the effluent is sufficiently purified. (2) Continuous * Winogradsky and Omeliansky, ‘Centralbl. f. Bakt.,’ 2 abt., 5, 1899. + Burri and Stutzer, ‘Centralbl. f. Bakt., 2 abt., 1 and 2, 1895 and 1896. } Stutzer and Hartleb, ‘Centralbl. f. Bakt.,’ 2 abt., 2 and 3, 1896 and 1897. § Fremlin, ‘Journal of Hygiene,’ vol. 3, 1903. || Winogradsky, ‘ Centralbl. f. Bakt.,’ 2 abt., 2, 1896. 4 Gartner, Frinkel, Kriiger, ‘Centralbl. f. Bakt.,’ 2 abt., 4, 1&98. ** Boulanger and Massol, ‘Ann. de l’Inst. Past.,’ vol. 17, 1903. tt Wimmer, ‘ Zeitschr. f. Hygiene,’ vol. 48, 1904. 1905. | with reference to the Purification of Sewage. 243 Filters.—In this procedure the liquor trickles continuously through the filter- bed, being uniformly distributed by sprinklers, while as perfect aération as possible of the bed is maintained. The objects of the present research were mainly the following :— 1. The detailed chemical study of the course of the nitrification occurring during the filtration of sewage, especially during the maturing period of the filter, and the comparison of the “contact ” and “continuous” methods (Section I). 2. The isolation and study of the organisms concerned, and comparison with those isolated from the soil by Winogradsky. The amount of organic matter accumulated in a sewage filter is comparatively great, and it seemed most unlikely that nitrification should here also be the work of bacteria so extremely sensitive to the presence of organic matter. One seemed com- pelled to believe that other and different bacteria must be here engaged (Section II). 3. The study of the question of absorption of ammonia upon the surface of filtering material previous to nitrification (Section III). These researches were begun in Vienna in 1901, and were resumed in Munich in 19035, after a break of two years. I am very happy to have this opportunity of thanking Professor Max Gruber for his kind hospitality extended to me in the hygienic institutes of both cities, as well as for the valuable advice and kind assistance he constantly gave me in the course of the work. My thanks are also due to the Royal Commission on Sewage Dis- posal for granting me leave of absence in 1903 to continue the research in Munich. I should also add that part of the expense of the work was defrayed by a grant from the Royal Society. Section 1—Chemical Study of Nitrijication in Experimental Filters. Description of Apparatus and Methods of Analysis—Small experimental filters were erected, consisting of glass cylinders 50 em. high and 12 em. in diameter ; these were placed one above the other, fitted well together by means of specially ground rims, and covered on the outside with black glazed paper. There were altogether three filters, differing only in height— 200 cm., 100 cm., and 50 cm. respectively. Fig. 1 is a diagram of the filter of medium height, showing the arrangements made to allow of samples being drawn off, and of the temperatures being measured at different depths; the tall filter, consisting of four cylinders, had the three upper ones similarly constructed. The filters were filled with small coke, carefully sifted and ot a uniform size (mean diameter 3-5 mm.). By volumetric measurements with 244 Dr. H. Chick. The Process of Nitrification | Apr. 1, water, it was found that when this coke is packed into a space, the volume of Fic. 1—Diagram of the “ Con- tinuous” Filter, of medium height, showing construction. The contour of the filter is represented as interrupted at K, to show the arrangement of the thermometer and the collecting funnel. the interspaces between the pieces is 35 per cent. ; the volume of the pores inside the pieces (amount of water retained on draining) is 20 per cent.; and the volume of solid coke is 45 per cent. of the whole space. ‘The three filters were fixed to the wall, near together, and all treated in exactly the same way, 7.¢., as continuous filters. Each received 4 litres of liquid daily, and the sewage employed was the liquid manure (“Jauche”) from a neighbouring cowshed; this proved a very suitable material, after a rough filtration through glass wool, and dilution to 1 in 20 with tap-water. This liquid was contained in a 10-litre reservoir bottle of the Mariotte type, from which it” dropped regularly into a small vessel containing a siphon arranged to empty when 100 c.c. had collected (A, fig. 1) into a sprinkler (B, fig. 1), the object of which was to distribute the liquid as evenly as possible over the surface of the filter. The methods of analysis employed in following the course of the oxidation of nitrogen were those usually adopted in such work,* but the following details may be given :— In the estimation of free and saline ammonia, 1 to 10 e.c. of the liquid was taken and diluted in a retort with about 500 c.c. NH;-free water, and distilled, it being found unnecessary to add any alkali Three successive portions of 50 ¢.¢. were distilled off and the ammonia they contained estimated by means of Nessler’s reagent. Albuminoid ammonia was afterwards estimated by adding to the same retort a definite amount of “alkaline permanganate solution,” distilling as long as ammonia came over in the distillate, and estimating these amounts in the same way. Oxidised nitrogen in the filtrates was detected by means of the reaction with diphenylamine sulphuric acid. Nitrites were distinguished by reactions with acidified * Cf. ‘ Report of Royal Commission on Sewage Disposal,’ vol. 4, part 5. + ‘Sew. Com. Report,’ vol. 4, part 5, appendices 3 and 4. 1905. | with reference to the Purification of Sewage. 245 starch-zine-iodide solution, and with metaphenylene-diamine, and estimated by the use of the latter. When both nitrites and nitrates were present, they were estimated together by the indigo method (Tiemann-Giirtner’s “ Wasseranalysec”) and the figure for nitrates obtained by subtraction. This method was afterwards given up in favour of the copper- zine couple method,* where, to allow for. traces of ammonia originally present in the solution, or introduced during analysis, a control estimation was ulways made ; this control was carried out in every way like the real analysis, except that no couple was introduced, and the ammonia obtained was subtracted from that found in the actual estimation. Total nitrogen was estimated in the sewage by Kjeldahl’s method, a small amount of Na,SO, only being added during the preliminary heating. The oxidisability (‘‘ Oxidirbarkeit,” or measure of oxidisable substances present) was estimated by reduction of permanganate in alkaline solution on boiling for ten minutes, care being taken to keep the external conditions (such as concentration, size of flask, total amount of liquid present, temperature, and time of reaction) constant in all determina- tions. Periodic examination of sewage and filtrate in this way gave useful comparative results and a means of following the course of the general oxidation. Cowrse of General Oxidation During Maturation: First Appearance of Oxidised Nitvogen.—Vhe filters were started on February 20, 1901, and their action carefully controlled by means of analysis from that time onwards. Especial attention was paid to the period of maturation from the time of first using to that of full efficiency, as it was thought this should throw light generally upon the manner of their working. The sewage employed contained on an average :—total nitrogen 10 parts, organic matter (by evaporation and ignition) 21 parts; oxidisability (expressed in terms of O) about 11 parts per 100,000 by weight. The course of the oxidation will be seen by reference to Tables Ia, Ip and II. On March 7 (see Table II, analyses 1 to +) there was a marked amount of general oxidation taking place in all three filters, or at least a reduction of oxidisable substances, but there was no trace of oxidised nitrogen in the filtrates nor was there any diminution in ammonia. Ovxidised nitrogen first appeared.in the tall filter on March 18 (the filtrate being then clear, bright and without smell), three days later it was detected in the medium filter and seven days later in the short filter. During this first period (after starting and before the occurrence of nitrification) the amount of ammonia in the sewage was frequently compared with that in the filtrates and found to be always the same.f On March 7 this was proved to be so for all three filtrates, and even on March 28, when the first trace of oxidised nitrogen appeared in the filtrate from the short filter there was no diminution of the free and saline ammonia coming through (¢f. analyses 10 and 13). So, generally, Tables II and IIT * Sutton, ‘ Vol. Analysis,’ 8th ed., p. 452. + As the sewage varied considerably in composition from time to time, care was always taken that the sewage and filtrate analysed for comparison should correspond to one another as nearly as possible. i 246 Dr. H. Chick. The Process of Nitrification | Apr. 1, seem to show that loss of ammonia goes hand in hand with production of oxidised nitrogen. Tables IA (Vienna) and Isp (Munich) give a clear idea of the general progress of the oxidation of nitrogen. They are compiled from analyses made from time to time, and while the times given for the later stages are approximate, those for the first stage are exact, being based on almost daily tests. The Nitrite Stage—After the first appearance of oxidised nitrogen in the filtrates, nitrification went ahead, and, in the case of the tall filter, five days later there was nitrous nitrogen in the filtrate about equal in amount to the total nitrogen going on. The two stages in which nitrification occurs were well separated in time, and show very distinctly, first, the production of nitrites in quantity without nitrates, and finally the complete oxidation to nitrates, nitrites being absent. For example, in the case of the tall filter, five weeks after having been started (Table II, analyses 10 and 11) the sewage contained 17 parts ammoniacal and albuminoid nitrogen, which in the filtrate was reduced to 1:5 parts, while 10 parts nitrogen were present as nitrites, nitrates being altogether absent. Three days later a similar result was obtained (analyses 14 and 15), eight weeks later nitrates were being formed in small amount, and an analysis of the filtrate made four months after starting (analysis 21) showed a complete oxidation of the nitrogen, nitrates being present in quantity unaccompanied by nitrites. In the case of the short filter the process was much slower, for an analysis made after four months showed production in the filtrates of nitrites only (analysis 23). This comparative lack of efficiency may be referred to the lower temperature* of the short filter, for an analysis on the same date of the liquid from No. 1 tap of the tall filter, 50 cm. from the top (Table III, 6 and 7) showed the presence of nitrates in abundance, and only traces of nitrites. * An exactly parallel fact was noticed with regard to the time of the first appearance of oxidised nitrogen in these two filters. Strictly speaking, the top section of the tall filter should have been exactly comparable with the short filter, and the only possible explanation is that, as the first division of the latter filter stood 150 cm. higher in the room, the discrepancy was due to a temperature difference. Regular observations of temperature had been made by means of the thermometers in the filters, and the higher position in the room was found to be constantly from 1°5° to 2°5° C. higher than the lower one, the temperature of the interior of the filters differing hardly at all from that of the surrounding air. This explanation is confirmed by the results of the Munich experiment, where all the six filters were arranged so that their tops, and not their bottoms, were on one level ; hence the first divisions of all were at a similar temperature, and it was found that the first oxidation of nitrogen was observed after the same period in all the three filters of each set. 1905. | with reference to the Purification of Sewage. 247 This complete separation of a nitrite from a nitrate stage is doubtless due to the comparatively strongly ammoniacal nature of the sewage employed. Previous observers*f have shown the inhibitive effect of ammonia upon nitrate-production, and it is probable that during the earlier stages of the maturing period the nitrate bacteria were unable to become established in the filter, and only later, when the ammonia of the sewage was being rapidly oxidised to nitrites, was the environment suited to their growth and develop- ment. The sewage employed frequently contained more than 15 parts free and saline ammonia per 100,000, a concentration which has been shownf to be sufficiently high to check completely the production of nitrates in pure culture. An interesting confirmation of this explanation was obtained in the maturing of the Munich filters, where nitrates appeared in the filtrates very soon after the first appearance of nitrites. There was here no such “ nitrite stage,” and the sewage was much less ammoniacal (2 to 4 parts ammoniacal nitrogen per 100,000). Difference of Function in Different Strata of the Filters—An attempt was made to study the course of the oxidation at different depths in the tall and medium filters (see Table III). In analysis 3 the “ oxidisability ” was taken as a criterion and the decrease in the first 50 cm. of the tall filter was found to be almost as great as in the whole length of the filter. One may therefore suppose that the mechanical deposition of suspended particles as well as the absorption of the more complicated organic matter in solution takes place principally in the upper layers of the filter. It is also apparent from analysis 4 that the formation of nitrites (these analyses were made during the nitrite stage) did not at that date take place in quantity in this upper layer but lower down for the most part; this same fact is also shown in analysis 2. In the latter case the free and saline ammonia was also estimated, and the decrease, which is so marked as the sewage passes through the filter, was found not to begin until after the first 50 cm. were passed. The same phenomenon appears in analysis 8 and is a striking instance of the principle, already alluded to, and discussed at length in the section devoted to absorption, that the dis- appearance of ammonia and the oxidation of nitrogen are closely associated both in time and space. Comparison of Contact and Continuous Filters—Munich Experiments — From the mature Vienna filters attempts were made to isolate the nitrifying organisms, but before much progress had been made the work was dis- continued and was not again resumed until after two years. This second time, in Munich, in 1903, fresh filters had to be matured, and a second * See footnote **, p. 242. + Warington, ‘Chem. Soc. Journ.,’ vol. 35, 1879, and vol. 59, 1891. 248 Dr. H. Chick. The Process of Nitrification [Apr. 1, opportunity was afforded for studying the maturing process, and nitrification generally. Filters like those previously described were again erected, and three of them treated as formerly with a continuous trickle of diluted liquid manure. Another similar set of three filters was treated, for contrast, as contact filters by the procedure mentioned on p. 242. Eight litres of sewage were treated every 48 hours by each of the filters. The capacities of the tall,* medium and short filters were respectively, after wetting, 6, 4 and 2 litres, so that each filling remained in contact at least four hours in the two taller filters and two hours in the short one. For 38 out of the 48 hours of the cycle the filters were empty oremptying. ‘These contact filters were cone-shaped below, with a narrow opening that could be closed with a cork for the purpose of filling them. In the case of the continuous filters, oxidised nitrogen again made its first appearance in the filtrates four weeks after starting (compare Tables JA and 1s). The fact that all three Munich filters behaved alike in this respect (forming a contrast to the Vienna ones) has already been explained as a temperature effect, see p. 246. It was noticed that during the maturing period the two different stages of nitrogen oxidation merged one into the other, and were not so clearly separated as was the case in the Vienna filters (Tables LA and Is); noticeably there was here no long period in which nitrites were formed in quantity without any accompanying nitrates. This difference has already been discussed (p. 247), and explanation is doubtless to be found in the much less ammoniacal nature of the sewage here employed. The contact filters did not yield nearly such good results as the continuous filters (Table Is). The period which elapsed before nitrogen oxidation was apparent was, in the former, more than half as long again as in the latter. Again, when the short continuous filter showed complete oxidation of its nitrogen, the tall contact filter still showed presence of nitrites im its filtrate. — The Munich continnous filters had completely matured in about ten weeks from the time of starting, but they were yielding a very satisfactory effluent much earlier. After three months the sewage was changed for a much more strongly ammoniacal liquid (cows’ urine, diluted 1 in 100, containing 14 to 17 parts ammoniacal nitrogen per 100,000), in order to test the capabilities of the filters as regards nitrogen oxidation. Most satisfactory results were obtained (Table II, analyses 33 to 56, and Table IV), the filtrates contained, as a rule, only traces of ammonia and nitrites, but abundance of nitrates. Attempt to further tax the capabilities of the filters * Tn these Munich experiments the tall filters, both contact and SERENE OS, were only 150 em. high, instead of 200 cm., as in Vienna. 1905. | with reference to the Purification of Sewage. 249 met with failure. Cows’ urine diluted only 1 in 50, and containing about 30 parts ammoniacal nitrogen per 100,000, was put through the medium filter for about a fortnight, and also at a later date through all three filters. but it was found that they were incapable of oxidising so concentrated a liquid, and the quality of the filtrates deteriorated (¢7: Table IV). Throughout the history of these filters there was a considerable loss of total nitrogen from the sewage while filtering through, but it was specially noticeable during the period when the diluted urine was being treated, when in some cases not much more than half the original nitrogen was present in the filtrate (Table II, analyses 33 to 56). This loss is doubtless due to an escape of free nitrogen, set free possibly by decomposition of ammonium nitrite, a very probable intermediate product in the nitrification of ammonia (NH,NO, = 2H20 + N2). This loss of nitrogen was not so marked in the ease of the Vienna filters (Table II, analyses 1 to 25), though it occurred later to some extent. These differences are probably due to the absence or presence in quantity of the organisms involved. The Munich continuous filters in their later history, and they were worked for about a year, possessed an efficiency rarely met with in large scale filters, showing that this type can give excellent results in the absence of much suspended matter.* The larger filters could not be considered to be heavily worked, but the short filter, which had a capacity of two litres and treated four litres of sewage daily, approximated more nearly to a practical installa- tion. It oxidised daily about 0°5 gramme nitrogen, and this resuit must be considered extremely satisfactory when the high nitrogenous concentration of this special sewage (14 to 20 parts of ammoniacal nitrogen per 100,000 instead of the 3 to 8 parts usual in ordinary sewage) is kept in mind. The quantities of nitrate appearing in the filtrates from these filters have rarely, if ever, been obtained in practice on the large scale. Section 11.—Bauciteriological Investigations. Enumeration of Nurifying Bacteria in the Filtrates—It was thought worth while to attempt to count the numbers of nitrifying bacteriat present in the filtrates from the filters, as it was conceivable that such enumerations might furnish a bacteriological criterion of the quality of sewage effluents. The practical utility of this procedure is, however, diminished by its * This can be removed in practice by a preliminary screening or septic tank treatment. + The recent work of C. C. Frye (‘Report of Roy. Com. on- Sewage Disposal,’ vol. 2, 1902, p. 9) has experimentally verified the view which has been generally held, though doubted in some quarters, that all the nitrification taking place in sewage filters is the work of hving organisms, and none of it purely chemical. 250 Dr. H. Chick. The Process of Nitrification [Apr. 1, slowness, due to the sluggish growth of the organisms and to the extremely small numbers of them introduced in the higher dilutions. The method might, perhaps, give useful comparative results even without allowing the maximum time for development in the subcultures, and the rate of growth could be accelerated by a temperature of 28° to 30°C. The enumerations were made as follows :—The filtrate was successively diluted with sterile water to a tenth degree six times ; of these six dilutions (viz., 1 in 10, 1 in 100, 1 in 1000, 1 in 10,000, 1 in 100,000, 1 im 1,000,000) 1 ¢.c. was used in every case for inoculation into bouillon and into Winogradsky’s ammonia and nitrite-containing media respectively,* which were distributed in test-tubes, each containing about 5 ¢.c. The ammonia tubes were subsequently tested for production of nitrites with acidified starch- zinc-iodide solution, and the nitrite tubes for nitrates with diphenylamine sulphuric acid (after evaporation to dryness with NH,Cl if any nitrite remained unoxidised). The numbers in which the nitrifying bacteria are present are surprisingly large (see Table V), and it will be seen that there is no strict relation between the numbers present respectively of nitrite and nitrate producers ; but the latter would appear to be present generally in less amount even when the filtrate shows complete oxidation of its nitrogen to nitrates. The filtrates used were in all cases from the Munich continuous filters. Isolation of the Nitrite Producer—tThe isolation of a nitrite-producing bacterium in pure culture was found to present considerable difficulty and many unsuccessful trials were made. Attempts were first made to isolate it directly from the coke of the filters by the method of dilutions. This method, originally invented by Lister, was formerly employed by Waringtont and P. and G. Franklandt for the same purpose, but with only partial success. The method here employed was similar to that used by the Franklands, except that much higher dilutions were made, and a larger number of tubes (about 200) containing appropriate culture media§ were sown with small amounts of liquid from the higher dilutions, bouillon tubes being also similarly inoculated as controls. Repeated attempts to isolate from the coke of the filter were only partially successful.|| A culture was, however, obtained which was comparatively, but not absolutely, pure ; this was used for further isolation experiments, and will be referred to as eulture “a.” * Omeliansky, ‘Centralbl. f. Bakt.,’ 2 abt., 5, 1899. + See footnote, p. 247. { P. and G. Frankland, ‘ Phil. Trans.,’ B, vol. 181, 1890. § Culture solutions used were diluted urine and Winogradsky’s solutions. || Starting from such very impure material, the dilution method does not give an adequate return for the great labour it entails. The filtrates might have proved better original material, as in them the nitrifiers sometimes predominated (see Table Y). 1905. | with reference to the Purification of Sewage. 251 Culture “a,” and all cultures showing production of nitrites* invariably contained in quantity a small oval bacillus or coccus, which was recognised as the nitrite-producing organism. Attempts to obtain a pure culture were further made with the use of ammonium agar as medium,t but without success. The plate cultures showed vigorous formation of nitrites,* but all nitrifying subcultures were found to be impure. The employment of a similar medium composed of agar and diluted cows’ urine was equally unsuccessful. Ordinary gelatine plate cultures were made and bouillon was inoculated from the impure cultures; none of the colonies separated from the former were able to nitrify, although 40 were investigated. From the growths in bouillon, plate cultures were also made on nutrient gelatine and agar, and 70 of the organisms separated were further investigated, but in no ease did nitrification occur. This seemed to show that the nitrifying organisms in filters resembled those of Winogradsky very closely. Therefore, in order to decide if the nitrifying organism was or was not able to live in the bouillon, an ammoniacal medium was directly inoculated from the growths in bouillon. Usually there was no nitrite-production (¢g., Table VI, culture “a”), and indeed the oval bacillus could in no case be traced in the bouillon growths. In one instance, however (Table VI, culture “d”), inoculation from a bouillon growth led to nitrification, but this property was lost after a second generation in bouillon; it therefore seemed probable that, if the nitrifier had not been killed in the bouillon, it certainly had not been able to multiply there. It thus was evident that, contrary to expectation, the nitrite-producing organisms of sewage filters were also unable to grow upon media containing organic matter; recourse was then had to silica plate cultures, which were made and inoculated according to the directions given by Omeliansky.t This operation was accomplished with comparative ease if the original sodium silicate was quite pure; the study and isolation of the separate colonies was, however, found to be exceedingly difficult. The sub-cultures * The test for production of nitrites was usually made by allowing a little of the culture fluid, withdrawn with a sterile pipette, to drop into a small quantity of acidified starch-zinc-iodide in a porcelain dish. This was preferred to the similar test with diphenylamine, partly because of its specific nature, and partly because the ferric salt present in the sediment of the culture-tubes also yielded a slight blue colour with diphenylamine. + (NH4)2S0,, 2:0 gr. ; NaCl, 20 gr. ; K2HPO,, 1:0 gr. ; MgSO,, 0°5 gr. ; MgCO,, in excess ; agar-agar (purified by washing, Beyerinck’s method, ‘ Centralbl. f. Bakt.,’ vol. 19, 1896), 20 gr. ; distilled water, 1 litre. { See footnote, p. 250. VOL. LXXVII.—B. T 252 Dr. H. Chick. The Process of Nitrification [Apr. 1, obtained were to all intents and purposes pure cultures, showing pure pictures of an oval, almost spherical organism, resembling the nitrosomonas of Winogradsky, except that it seemed to be somewhat smaller in size. It appeared constantly in the form of zoogloea embedded in the particles of magnesium carbonate at the bottom of the culture tubes, and it stained easily and well. The individual bacteria were often found to be well separated in a culture, but an actively motile stage was not observed. These cultures, however, still gave a growth, though extremely slow, in bouillon, and this consisted of the other quite inconspicuous organisms present. By means of the dilution method, pure cultures were obtained which yielded absolutely no growth in bouillon when preserved indefinitely either at 37° or at the room temperature. These pure cultures were not, however, robust, and they nitrified very feebly; attempts are now being made to obtain vigorous pure cultures. Isolation of the Nitrate-producer.—The dilution method was also employed for the isolation of the nitrate organism, the original material being a culture obtained during the enumeration experiments (Table V), which showed active oxidation of nitrites. A culture was separated which consisted of the nitrate bacterlum mixed with one other species, and the combination, referred to in future as culture “d,’ formed a very interesting symbiosis. Pure cultures were obtained from culture “d” by making surface plate cultures, in great dilution, on nitrite agar.* These pure cultures showed a small non-motile bacterium, agreeing in essentials with Winogradsky’s organism, though somewhat larger in size. It was a bacterium very thick in comparison with its length, so that it often appeared to be almost a coccus; stains were badly taken up, and it frequently appeared imperfectly and irregularly stained. These pure cultures rapidly changed nitrite to nitrate, when growing in nitrite-containing medium,} the nitrite present being sqmetimes completely oxidised in less than two weeks. Bouillon on the other hand remained indefinitely sterile; the tubes were kept under observation for seven weeks without there being any sign of growth. * Omeliansky, ‘ Centralbl. f. Bakt.,’ 2 abt. 5, 1899. + Winogradsky’s nitrite culture solution was invariably employed, and the cultures were tested from time to time for the production of nitrates. When time enough had elapsed and all nitrite had disappeared, then, on testing the culture liquid, a negative reaction with starch-zinc-iodide, and a positive with diphenylamine proved the presence of nitrates. But if all nitrite were not oxidised, the remainder was decomposed by evaporating to dryness with a little NH,Cl, and the residue dissolved in water and tested for nitrates with diphenylamine. This method has been shown to be quantitative when such substances as sugar and peptone are present (Frankland, ‘ Journ. Chem. Soc.,’ 1888), and it is possible it might also prove a useful method of estimating nitrates in presence of nitrites in sewage effluents. 1905. | with reference to the Purification of Sewage. 253 Pure cultures were also obtained with more difficulty directly from less pure material, by means of nitrite agar plates, but the organism isolated was in every case the same. All attempts to isolate a nitrate organism by means of ordinary nutrient agar and gelatine were unsuccessful. In no instance was nitrite oxidised to nitrate by any organism separated on such plate cultures, though over 40 such organisms were investigated. Experiments with “ Symbiotic” Cultures of the Nitrate-producer.—Although nitrobacter, when alone, is incapable of growing in bouillon, it would appear to be capable of surviving an inoculation into bouillon if not alone, but growing with certain other bacteria. A very instructive set of experiments was made with culture “d” (Table VII), in which this strain was inoculated into bouillon through four generations. From each set of tubes nitrite medium was inoculated, and it was found that the change to nitrate occurred invariably in the tubes sown from the earliest bouillon generation, and in two instances also from those sown from bouillon of the fourth generation (Table VII, d3 and ds). The quantities inoculated were large, one or two drops, but it is impossible to believe that nitrobacter would still be present in a fourth generation if no multiplication had taken place in the bouillon. Pure cultures of the uitrate-producer showed no such effects ; inoculated bouillon remained quite clear; examined under the microscope it showed complete absence of bacteria, and nitrite tubes inoculated from the bouillon in no case showed any oxidation to nitrate. In Table VIII are shown the results of further experiments in which three pure and four mixed cultures were compared in this respect, and one is compelled to conclude, in explanation, that the presence of the accompanying organism in some way protects the nitrate bacterium from adverse influences present in the bouillon, which it is unable to withstand if alone. Without further experiment, any attempt to explain in what this action really consists must be pure conjecture, but it is possible that the harmful organic substances present are in some way altered by the accompanying organism, and it would be interesting to see whether the nitrate organism in pure culture could thrive in bouillon previously exhausted by its companion.* Phenomena which present an interesting analogy with these observations are found in the case of certain anaerobic organisms, one instance of which has been precisely investigated by Winogradsky, viz. that of Clostridium Pasteurianum.t This strictly anaerobie species was found to be capable of * The experiment was not made in this instance because the cultures had then been isolated a considerable time and their properties were enfeebled. + ‘ Archives des Sciences biolog. de St. Petersb.,’ vol. 3, 1895. 254 Dr. H. Chick. The Process of Nitrification [Apr. 1, growing aerobically when, and only when, associated symbiotically with a certain aerobic organism which removed the surrounding oxygen and created an oxygen-free environment for it. Such symbioses of various grades must be frequent in Nature where the “pure culture” is almost unknown. The part played by the artificial pure culture in the progress of bacteriology has, of course, been enormous, yet its possibilities are limited, and one must look to the investigation of regulated simple symbioses for a nearer approach, in the laboratory, to the workings of Nature. A break of two years occurred during the course of these investigations. After they were again resumed, Dr. Schultz-Schultzenstein* published the results of bacteriological investigations, having the same aim as the present work. He isolated two kinds of nitrifying organisms from the material of coke sewage filters at Karolinenhéhe, near Charlottenburg, which corresponded exactly to those isolated from the soil by Winogradsky, and no other nitrifying organisms were found. His researches must be regarded as the first published successful attempt to investigate the organisms concerned with nitrification during the artificial purification of sewage, and the results are entirely confirmed by the present investigation. In spite of this anticipation of my identification of these bacteria, I have thought it worth while to describe my isolation experiments in detail, because in a subject of such technical difficulty the experience of an independent worker may be of use to others. Section III.—Absorption of Ammonia and Ammoniacal Compounds during Sewage Purification. It has been held that a most important preliminary to nitrification, both in the soil and in sewage filters, is to be found in an absorption of ammonia and ammonium compounds upon the surface of the particles of soil or of filtering material respectively. In the case of the soil, a long controversy has taken place as to whether a physical or a chemical process was here in question, and the former view, maintained notably by Liebigt and his school, has on the whole prevailed. This “adsorption” of ammonia plays an important part in the current doctrine of the action of sewage filters, which considers that nitrification could not take place in the short time taken by * Schultz-Schultzenstein, ‘Mitt. a. d. Kon.-Priifungsanst. f. Wasservers. u. Abwas- serbeseit,’ 1903. + Way, ‘Agric. Soc. England Journ.,’ series 1, vols. 11—13, 1850—1852, and Mayer, ‘Lehrbuch der Agrikulturchemie,’ 1871. Lemberg, ‘Zeitschr. d. deutsch. geol. Gesellsch.,’ vol. 28, 1876. {t ‘Liebig, ‘Ann. Chem. Pharm.,’ vol. 94, 1855, aud vols. 105 and 106, 1858. 1905. | with reference to the Purification of Sewage. 255 the liquid to pass through the filter, and that the nitrites and nitrates appearing at any particular time in the filtrates are the result of a slower change which has been effected by the nitrifying bacteria upon ammonia previously absorbed in some physical manner upon the surface of the filtering material. The procedure for the purification of sewage used in “contact beds” has been held to assist successively the processes of adsorption of ammonia and nitrification. Dunbar and Thumm* consider that, in the “filling” and “full” stages, putrescible and oxidisable substances are retained upon the surface of the filtering material, and are subsequently oxidised at times when the bed is full of air, the oxidation being the work of bacteria in the bed, among which the nitrifying bacteria rank high in order of importance. As regards the complex oxidisable putrescible substances of high molecular weight, the solid suspended matter will be retained, of course, by mechanical filtration, while the soluble constituents may, doubtless, be supposed to undergo some physical adsorption.ti But the greater part of the nitrogen present in sewage is there in the form of free and saline ammonia, and these are the compounds most markedly retained as the sewage passes through the filter; yet for such simple compounds as these, adsorption by solids has been shown to take place only to a small degree or not at all§ To attempt to explain removal of ammonia by adsorption then, would appear inadequate. Special experiments were therefore made to investigate the behaviour of filtering materials with ammonia and its salts ; also during the investigations in Section I, careful note was also made of any facts which should tend to confirm or refute the theory of nitrification quoted above. If the theory of a previous ammonia absorption and a subsequent oxidation were true, then contact beds should be much more efficient nitrifiers than continuous filters, but the contrary proved to be the case, the latter doubtless owing their greater efficiency to their more perfect aération. Again, while complicated organic substances appeared to be absorbed in the top layer of the filter (Table III, 3), the disappearance of free and saline ammonia was shown usually to take place lower down in the filter(Table IIL, 1), and to be always associated with the appearance of oxidised nitrogen (Table III, 2, etc.). Moreover, during the maturing of the filters, before oxidation of nitrogen had occurred, no absorption of ammonia could be detected, although this was * Dunbar and Thumm, ‘ Beit. zur Abwasserreinigungsfrage,’ 1902. j Soyka, ‘ Archiv f. Hygiene,’ vol. 2, 1884. t Kattein and Liibbert, ‘Gesundheitsingenieur,’ vol. 25, 1903. § Weppen, ‘Ann. d. Chem. u. Pharm.,’ vol. 55, 1845, and A. Mayer, ‘Lehrbuch d Agrikulturchemie,’ 1871. 256 Dr. H. Chick. The Process of Nitrification [Apr. 1, frequently looked for. It was possible, however, that, at the very first, ammonia had been taken up by the filtering material, to saturation point, and that afterwards no more absorption was possible until nitrification had begun. Unfortunately no analyses were made at the very beginning, but this gap was afterwards filled by special experiments with clean sterile coke. The tendency of these observations was thus in opposition to any theory of ammonia-absorption by the filtering material, and this opposition was con- firmed by the following experiments, made to test the power of various solids to absorb ammonium salts. The solids employed were barium sulphate, sand, and ground-up “clinker,’ and the experiments were carried out as follows :— A small quantity of a solid (1 gramme or 2 grammes), previously carefully purified, was weighed out into a small flask, which was then exhausted to remove air films, which might cause imperfect contact of solid and liquid.* A measured quantity (50 c.c. to 100 c.c.) of ammonium chloride solution was added through a tap-funnel, and the whole left standing for 24 hours. The clear liquid was then drawn off by a pipette, and the ammonia estimated in a small portion (1 to 5 ¢.c.); this was first diluted to about 500 ¢c.c. with NH,-free water, and then distilled and nesslerised. The remaining liquid was well shaken up and the muddy residue analysed similarly : a correction had to be made for the volume of the solid, which was measured, after centrifugalisation, in a graduated tube. In every case, as a control, a blank experiment was also made, similar in every detail except that NH,-free water replaced the ammonium chloride solution. The solids had previously undergone a careful purification by washing, and often, by ignition also. The ammonium chloride solutions were exceedingly dilute, so as to approxi- mate to the concentration of ammonia in ordinary sewage. In Experiments 1 to 3 (Table [X) the ammonia yielded, both by the clear and the muddy portions of the liquid, was found to have diminished. It was therefore supposed that boiling was insufficient to drive off any ammonia which might have been absorbed by the solid; accordingly, in Experiments 4 to 7, a small amount (10 cc. N/1 KOH) of alkali was added before distil- lation. In this case the analysis of the muddy portion of the liquid showed a small amount of the ammonium salt to have been absorbed by the solid, but nothing comparable to the effect required in a sewage filter. Moreover, the slight removal of ammonia demonstrated would appear to be a chemical rather than a physical phenomenon, alkali being necessary to free the absorbed ammonia from the solid. In none of these experiments, however, was coke itself employed, and the surface of solid was very small in comparison with the amount of liquid taken ; * In Experiments 6 and 7 (Table IX) the flask was not thus exhausted, and the agree- ment of their results with those of previous experiments indicates that this precaution is unnecessary. 1905. | with reference to the Purification of Sewage. 257 the following further experiments were therefore made, and confirmed the preceding ones. Experiment 8.—Exactly the same coke as had been used for the filters was taken and thoroughly washed and dried. An amount occupying a volume of 30 c.c. was placed in a flask with 50 c.c. NH,Cl solution of concentration equal to about 5 parts ammonia per 100,000. In a control flask 50 ¢.c. of the liquid was placed alone. After 24 hours and after 48 hours the liquids in the two cases were examined, a small quantity (0°5 c.c. to 1 ¢.c.) being removed, diluted to 50 c.c. with NH;-free water and tested with Nessler’s reagent. In no case was the reaction fainter where the liquid had been in contact with the coke. (It was shown that the coke did not of itself yield ammonia by a control experiment in which NH;-free water replaced the NH,Cl solution.) Experiment 9.—Coke, which had been thoroughly washed, dried, and sterilised, was placed in a cylinder to form a small filter, and a solution of NH,Cl (10 parts NH, per 100,000) allowed to drop slowly through.* The filter occupied a volume of about 1 litre and during the first hour 50 c.c. came through, while in 15 hours a total of 500 c.c. was filtered. The first filtrate of 50 c.c. was tested for ammonia and compared with the original liquid. The tint given by the filtrate (after suitable dilution and addition of Nessler’s reagent) was, if at all, only a shade paler, indicating only a negligible difference. After a second hour the filtrate was again compared with the control, and a similar result was obtained. The filtrate coming through in the next 13 hours was similarly tested, but no absorption of ammonia was detected.t It may be objected that experiments with raw, cleansed, filtering material are not applicable to the occurrences in the mature filter, where the surface of the coke is probably coated, in some manner not yet investigated, and might possess the faculty of absorbing ammonia in a manner similar to that already demonstrated in the case of certain colloidal substances.{ Therefore it is hoped, in the future, to make experiments with matured coke, elimi- nating, if possible, the action of bacteria. The available evidence is, however, opposed te such absorption, for it is in the uppermost layers of the filter that such a coating would be greatest, and yet disappearance of ammonia, at any rate during the maturing period, has been shown to take place lower down, and in any case coinciding, both as regards time and place, with nitrification. Upon consideration of the experimental data at present available, one is therefore inclined to reject the current theory of nitrification and to consider * The concentration of ammonia was greater than in the preceding experiments, where it approximated to that in ordinary sewage; here a more concentrated liquid was employed, in order to be comparable with the diluted urine which was then being treated on the filters. + These last two experiments are in perfect accord with some of A. Mayer (‘ Lehrbuch d. Agrikulturchem.,’ 1871), who showed that pure carbon in a porous condition was unable to effect any significant absorption with many salts Jong known to be absorbed by the soil. t Van Bemmeden, ‘ Landw. Versuchsst.,’ vol. 35, 1888, ‘ Zeitschr. f. physikal. Chem., vol. 18. 258 Dr. H. Chick. The Process of Nitrification [Apr. 1, the disappearance and oxidation of the ammonia to be parts of one process, which is carried out by the nitrifying bacteria in the time taken by the sewage to pass through the filters. For these experimental filters the time taken for the passage was measured directly.* For the Vienna continuous filters on July 1, 1901, the time was approximately 34 hours for the tall and medium filters and only 5 minutes for the short filter (nitrogen oxidation in this filter had not then progressed beyond formation of nitrites). For the Munich continuous filters on March 17, 1903, the time was 2 hours for the tall filter, and $ hour for the short one. At this date, a too concentrated sewage was being employed, and the filters were not at their best, but still four-fifths of the ammonia in the sewage was oxidised while passing through the tall filter. Section LV.—General Conclusions. 1. Nitrification of ammonia during sewage purification occurs in two stages which may be referred to the activity of two classes of bacteria, one producing nitrites, and the second oxidising the nitrites to nitrates. These bacteria exist not only in the substance of the filter, but are also carried away in large quantities in the filtrates. 2. These organisms belong to the same group as those concerned with nitrification in the soil, isolated by Winogradsky. It is, at first, difficult to understand how organisms so susceptible to the presence of organic matter are able to live and do their work in sewage filters. The following, one or all, form possible explanations. (a) The nitrifying bacteria may be, to a certain extent, protected by the presence of other organisms, and this view is strengthened by the results of certain experiments with the nitrate-producer; in symbiosis with such organisms, made in the course of the present investigation. (6) It has been shown that porous materials, such as coke, are able to retain upon their surface complicated organic substances of high molecular weight, when these are presented in solution. We may suppose this absorption (together with the mechanical separation of the suspended * 100 cc. of a 2-per-cent. solution of sodium chloride were sprinkled over the top of the — filter (that being the volume of liquid usually delivered at each discharge of the siphon). The filtrates were then continuously tested with silver nitrate until a copious precipitation was obtained; the ordinary sewage filtrate yielded only a slight reaction with silver nitrate. The passage of liquids through such filters is a very complicated process and one not yet thoroughly investigated, and though, doubtless, the times thus determined may be considered to apply to the majority of the liquid going on at any particular time, they must still, strictly speaking, be regarded as approximate and indeed minimal values. 1905. | with reference to the Purification of Sewage. 259 materials in sewage, also largely of organic origin) to take place principally in the upper layers of the filter. The nitrifying organisms will then be able to live and multiply lower down in the filter where the amount of organic matter present will be comparatively small, and this view has been experi- mentally confirmed in the present work. (c) It has been lately shown by Wimmer,* in the case of the nitrate organism, that a porous medium has a markedly mitigating effect when organic matter is present, and the coke and other materials of which sewage filters are made, are selected mainly on account of their porosity. It is only fair, however, to state that Wimmer’s experiments were not made with absolutely pure cultures, and part of the beneficial effect observed may have been due to a symbiosis, though, from the nature of his experiments, it would seem unlikely. (d) The nitrifying bacteria are doubtless present in very great numbers in the filters, and this may assist them in withstanding the effect of organic matter. This view is based upon certain observations of Winogradsky and Omeliansky,* in which nitrifying organisms, if present in sufficient quantity, were shown to withstand amounts of organic matter otherwise inhibiting them. 3. In the maturing of sewage filters, the two stages of nitrification may be markedly separate in time (Vienna experiments), or may be both developed together (Munich experiments). This difference is correlated with the greater or less ammoniacal content of the sewage. In the stronger sewage used for the Vienna filters, the well known inhibitory action of abundance of ammoniacal compounds (especially of free ammonia and carbonate of ammonia, which are so largely represented in the sewage), presumably retarded the development of the nitrate-producer, until the nitrite-producer was sufticiently well established to be converting most of the ammonia into nitrites. 4. As a result of special experiments with coke, and of analyses of the filtrates at different depths of the filters, and at different stages during the maturing period, it would appear that there is no evidence of absorption of free and saline ammonia without contemporaneous nitrification. Further research is necessary, but the theory of a previous physical “adsorption ” of ammonia and subsequent slower nitrification would appear, at present, to be without experimental foundation. 5. One is therefore inclined, in the present state of our knowledge, to consider the process of nitrification, during the filtration of sewage through * See footnote, p. 242. + Lohnis, ‘Centralbl. f. Bakt.” 2 Abt., 13, 1904, and Boulanger and Massol, ‘Comptes Rendus, vol. 140, 1905. 260 Dr. H. Chick. The Process of Nitrification [Apr. 1, such filters, to be an extremely rapid biological process, requiring for its completion only the time taken for the liquid to pass through the filter (approximately 2 to 3 hours, possibly a little more). The rapidity of the process is probably to be explained by the very great number of nitrifying bacteria present and the very efficient aération which obtains. In such filters also, the general conditions are ideal for quick action, as the continuous trickle secures rapidity of diffusion, and forms a great contrast to the much slower effect in stationary fluids. 6. Temperature has a marked influence upon the oxidation of sewage, a higher temperature being noticeably more favourable. This indicates that the efficiency of sewage filters in practice would be much increased if at a reasonable cost they could be artificially maintained at a warm temperature during the winter.* 7. The previous conclusions are chiefly drawn from experiments with * continuous filters, but filters working as contact beds were also investigated and the two methods compared. On the “ammonia adsorption theory,” the contact method should have proved the most efficient. This, however, was not found to be the case. The advantages of the continuous method would seem to lie in the much more complete aération and efficient diffusion, and also in the stratified distribution in the filter of the different stages of the sewage purification. Some of the present experiments were quite comparable with practical installations as regards quantity of liquid treated, con- centration of nitrogen, etc., and the results were much more satisfactory than those usually obtained in practice. The obvious difficulty in practical employment of continuous filters is with regard to the solids in suspension, which can only be permitted upon the filter to a small extent without risk of clogging. The present experiments were all made with roughly filtered solutions, but the difficulty could be met in practice by a previous screening of the sewage or by passing it through a septic tank. Should clogging occur, it will probably take place in the superficial layers and could be remedied by simple mechanical treatment. In the case of contact beds, however, clogging necessitates the cleansing of the whole bed, an exceedingly costly process. From these considerations, and as a result of the present experimental study, the method of continuous filtration would appear to be a most advantageous method of purifying sewage. * Ducat filters, which are artificially warmed in cold weather, perform an amount of nitrification which is well above the average. 1905.] per 100,000 by weight. Tables I—IX. In these tables the estimations of “free and saline ammonia” are expressed as ammoniacal nitrogen, those of “albuminoid ammonia” as albuminoid nitrogen, those of nitrites as nztrous nitrogen, and those of nitrates as nitric nitrogen, all in parts of nitrogen with reference to the Purification of Sewage. 261 Nitrites were estimated by the metaphenylenediamine colorimetric method unless otherwise stated, the sum of nitrites and nitrates by the indigo method (Vienna analyses), or by the copper-zinc couple method (Munich analyses). Throughout these tables the grades of reactions are represented as follows: reaction absent by 0, a trace, or faint reaction, by f, definite reaction by +, and intense reaction by ++. Oxidisability is expressed in parts oxygen absorbed per 100,000 by weight, less that In figures marked with an asterisk the oxidisability absorbed by any nitrites present. was approximately gauged by the difference between the oxygen absorbed in the cold and on boiling. Table Is.—Vienna Continuous Filters, showing the Progress of Nitrogen Oxidation during Maturing. Filters all started on February 20, 1901. Figures in brackets are weeks elapsed since the start. Course of oxidation. Tall filter. Medium filter. Short filter. Oxidised N. first detected in| March 18 March 21 March 28 filtrate (4 w.) (4 w.) (5 w.) Nitrites present in quantity, no March 28* March 28+ July 1f nitrates (5 w.) (5 w.) (19 w.) Nitrates also present in quantity May 24 May 13 (11 w.) (94% w.) Nitrates alone present ............ July 1 July 1 | (19 w.) (19 w.) * 10 parts per 100,000 N. as N.O3. + 13 parts ditto. { 4:1 parts‘ditto, nitrates also present in small amount. Table Is.—Munich Experiments. Filters all started April 1, 1903. Continuous filters. | Contact filters. Course of oxidation. Tall. | Medium. | Short. Tall. Medium. | Short. | | Oxidised N. first de-| May2 | May 2 May 2 May 16 May 16 May 16 | tected in filtrate (4 w.) (4 w.) (4 w.) (63 w.) (63 w.) (6% w.) Nitrites present in| May 18 May 23 quantity, nitrates | (7 w.) (7% w.) present Majority of oxidised N.| May 27 — — | June 25 | June 25 | June 25 present as nitrates (8 w.) ~ (12 w.) (12 w.) (12 w.) Nitrates alone present.... June 8 June 13 | June 13 | July 9 July 9 July 9 | (0 w.) (103 w.) | (103 w.) | (143 w.) | (14h w.) | (142 w.) [Apr. 1, The Process of Nitrification Dr. H. Chick. 262 = 8-8 = T-8 0 41-0 30: 0 ceeeeees | Tye9—oq@dgT | 8O'L'ST 98 = I: 61 = 0) 0 L- 0- OT a 'd [fourm smo | ¢0'/'eT 9g 19. T g. 8 = £.8 0) SI-0 80: 0 "| TBI —oqeryL | g0'9'FS rE I OT 0: LT = 0 0 8-8 3: ST “9 'd T ‘our smo | ¢0'9%% 6 L-P 0-% = LT ) ¥- O 10-0 “" umipea— “| eg 9-8 8.3 = I. ) 2-0 (AON) asa Te 99849, heoguur-st 1g 6: OL 9. — (0) (0) 9.T ONee = aes “ oSeMos MBIT oe = 6-2 = 0-3 9-0 Z-0 -0 pee 11®}3— 998191], ua 6 aoe oe = i ‘ ae saa eee ned. GRUMG! den teo G' 12-98 { cs os 0. == G. 1 Fé TL z-0 1-0 unIpsu—ajyeuytT | OSes | LB = 0. a ) 0) ¥T 9-8 aoa oBUMOS MUY —-O'S"EZ 9 ard g.8 = ¥-0 G.% 3-0 ClOy ge i Sas 1%" S0'SET | Ge Q. Il I. G as, oO 0 Q. I Cc. @ seen eeeeenes asemos MOY 20'S'6T ind = = mr f LP = ERG gee PEE S10U8 ee LOWS £3 sige = = 0-g f = LF “ umipem— “| TOL'T ra = = = 8. L 0 = DOs Nae T133— 9984 = LOL 1z ae es ne 0 0 ae 6- TI eee eeeceseee aS'eMos MBIT | TO“ 0z = = = LT 0-9 = SHO see koes 13— =“ TO" "FZ 61 3G I-81 = = if 8-1 SHOR <1 |e qtoys— 81 «1-F g- 6 = L-& L- F-0 €-0 “+ UNIpeni—oqecy [Ly \o'g'st-et LT Gc. ST BLL —_— (0) (0) Gg. ps 7 i ae OCDORGROD esemos Muy | 91 = G- OT = i) ¥- OL €-8 Sagres T®3—9F244LT | LOFT ST fame lL. 91 4 LI 0) 0 0: Zz fap FL een eeeerence aseMos MOY | 10'r'T PL 9-F = = = f = PeGiley 2 eee qtoys— “| 10°88 81 = = = 0) 9-1 = T- ST “* umipeu— “ | T0's'sz ‘ae T-¢ 9: IT = 0 I- OT 88. 0 Cole liane * T1@3—e4TLT 10°88 II 8: BP @- LT 3: AT ) 0 0:3 CoG we. See esemes MeY — 10'S'8Z OL ¥-G = = 0 0 = aie Sioa qtoys—ozeryLy 10'8ST 6 &. ST — — 0 0 = Pan leeeaearee: osumes Muy 10'S ST 8 9. 9 noes a 0 (0) = — | ~—_«jodaacop56 910 S— “cc 10'@’eL L 3: = = 0 ) = aay oe Woe TI} —998t4LT | TO'S'ST 9 e. CL eis aes (0) 0 nee: —~ —— “(| J0000G000 eee osemos Mey | 10'2'E1 G Pp. 9 Book = 0 0 a 2 me (ee 410 s— “ 1O'e'L , p- = = 0 0 = aaa J|° wmtpem— “ TO's'L g 6.¢ = = 0 0 = Cae Sees TIe3— eT | TO'eL z ©. OL aes = 0 0 Le DFA Jews n ee eeeeee aSemos MBI Toe" I ‘© Jo sunI0e4 ur *poye[noeo Tqeppely “ueS01410 ‘ues01}1U “uoSdo1q1U “u830141U “oq JO FUSIOP oy “ssf BUY ‘AqTIqesIpIxQ | Uesomy1U [eyo], | ‘Mesoaqra [ey07, OLAGINT SNLOT4I NT plourmmaly | [vovmomumy | “poskyeue [eLeyepy a JO ‘ON 000°00T 10d sqred UI pesserdxe oi sy[NseL OT, “EOP T poyeqs [[e ‘sieq[y Snonuryuoo yoruny oY} YIM [eop 9¢—Fz sesdjeuy {10'Z'0Z% pojtegs [[e S109} SNonUTZUOD vUUOIA EY} YIM Teop SZ—T sesd[euy ‘soqvIq[IA pue JUON_V esemog jo sosd[euy— J] 9]qeI, 1905. | with reference to the Purification of Sewage. 263 Table I1I.—Showing the Course of Oxidation at Successive Depths in Tall and Medium Filters (Vienna). | he Filtrate at successive depths. | | [= = — No. Date. Constituents. | Sewage. Tap No.1, Tap No.2, Tap No.3, | Tap No. 4, | | 50cm. 100 cm. 150 cm. | 200 cm. | ‘Pall filter. | | 1 24.5.01 | Ammoniacal N. eel 278) POF 2°27 0°39 0-29 2 24.5.01 | Oxidised N....... Vai XO ah + ++ +++ 3 27.5.01 | Oxidisability Oe) i oad 2 -6* SEOMS |) 2e38* 4 27.5.01 | Nitrites............ _ ye St No. 2x5 | No.2x7 5 1.7.01 | Ammoniacal N.} 11°9 2°5 0°2 0 -04 0°06 | 6 1.7.01 | Nitrites............ 0 ofa 0 0 0 7 1.7.01 | Nitrates ......... 0 = + + 7°8 | Medium filter. 8 | 17.01 |Ammoniacal N.| 11:9 | 81 | 4°12 9 | 1.7.01 | Nitrites............ | 0) if fe 10 | 17.01 | Nitrates ......... Liat ¥. 4°96 | In 2 the test was made with diphenylamine, in 3 the estimations are approximate. The high numbers in filtrate of 8 are due to deterioration in efficiency of the filter following partial clogging. + Nitrous N = 6:0, nitric N = 111. Table 1V.—Munich Continuous Filters from June 24, 1903, to November 27, 1903, using Cows’ Urine diluted 1 in 100; this contained 14—17 parts NHz per 100,000 and no oxidised nitrogen. In cases marked with an asterisk the urine was only diluted to 1 in 50. | Tall filter. Medium filter. | Short filter. | Slate ES) Nitrite!) Nitrate ee | Nitrite | Nitrate) NH, | Nitrite Nitrate per re- re- werere= "||P Sore. | er re- re- 100,000. | action. | action. 100,000. | action. | action. |100,000.| action. | action. | | | | | | | | 26.6.03 | | 0 ++ je (6) } ++ | 1 ifs ++ | 30.6.03| O72 | O ++ Ol O | ++ 2—3 fa ++ 9.7.03} 0-05 | Oe a Ee. | 2—3* + | +t 1—2 + ++ | 13.7.08 | | | } | } aa | +0-2 | Doe sear | eae SF POV NN joerc Sg | (ee fess |. | 23.7.03| 0:05 | O | ++ | 4—5* Sy renee (eee Ae 1—2 + ++ | | 24.10.03 | 0°05 0 ++ 0-05 0 | ++ On| 0) Pee _ 27.11.03 | 0-02 10) ++ 0-04 0 GEE 0-04 0 ++ | 20.104) 5:0* | Ff ++ |8—10¥| ++ | +4 15* + ++ The ammonia estimations by direct Nesslerisation are only approximate ; or inconsi derable). nitrites were tested for by acidified starch-zinc-iodide; nitrates by diphenylamine (only specific if nitrites are absent 264 Dr. H. Chick. The Process of Nitrification . [Apr. 1, Table V.—Enumeration of Organisms in Filtrates. No. of organisms per cubic centimetre. F Time sub- Date: 2) Mupiehicon: cultures were 2 Growing in Nitrite- Nitrate- Kept. bouillon. producer. producer. 6.6.03 Mall: Ce etosnencsescrs 10,000 10,000 100* 6 weeks 6.6.03 Medium ......... 1,000 10,000 100 Co, 6.6.03 | Short.........000. 100,000 10,000 100 65 12.6.03 Nl lee Sere eenonancr 10,000 10,000 10,000 4, months 12.6.03 Medium ......... 100,000 100,000 1,000 4 12.6.03 | Short...........00 100,000 10,000 100 ee ae NS1G.O37 4 Dall eranacccesteee 10,000 10,000 100 Former 4 m. TSIG!03) Ml Shortie cere 1,000 1,000 10 { Latter 6 w. 24.6.03 Stalls Goaenosaveooene 10,000 1,000 10,000 4, months 24.6.03 Medium ......... 10,000 1,000,000 10,000 yas 27603) | iiglla eee etecee: 1,000 = 10,000 A 27.6.03 | Medium ......... 10,000 = 10,000 is, * Culture containing 0°01 c.c. filtrate, used for dilution experiment on p. 252. Table VI.—Results of Inoculating two different Impure Nitrite-producer Cultures, “a” and “db,” into Ammoniacal Media, both directly and after passing through one or two generations in bouillon. ++ indicates an intense nitrite reaction. 3 2 5 “a,” three | “a,” twelve | “6,” three “6.” four SPeGs.|) Seibenw ar sneer oe. weeks old. days old. weeks old. | weeks old. i From original material ...... ++ = = ++] ++ apr A From Series 1 ...............05 — Fe | bt Pa | ae = 3 From original through bouil- 0 @) 0 ++] ++ ++ lon (Series 2) 6 From Series 2 again, through — _— = (0) 0) _ bouillon (Series 5) 1905. ] with reference to the Purification of Sewage. 265 Table VII.—Results of Inoculating into Nitrite Medium of Five Strains of a “Symbiotic” Culture of Nitrate-Producer, “d,” after Successive Generations in Bouillon. + indicates a positive, + + an intense reaction.} ce d 1 ” (77 d 2 ” | “cc d 3 ” ii3 d 4? | “ce d 5 ” Dates of Date of Date of successive | inoculation aoe g | inoculation from bouillon eerie ale eo ee ee ie eee into into nitrite | Preceding | $/8/28)/2/2/28|/2)/2|28)3 bouillon. medium. cultures: lee Boe rer |B |e | 8 | & £ | 17.12.03 -» 23.12.03 12.1.04 O f+ +! O }++ ++ 12.104 + 18.1.04 3.2.04 oy) | 0 leans O |++|) O j++} O |+4 18.1.04 > 26.1.04 10.3.04 ++)| 0 (0) aaa O |}++| O |++/+-+4) 0 26.1.04 > 3.2.04 HOBO, Wes | Ogee! OP ee ca Table VIII.—Showing the Different Behaviour of Pure and of “Symbiotic” Bouillon-Cultures of the Nitrate-Producer when Inoculated into Nitrite Medium, the latter tested for nitrites and nitrates 31 days after inoculation ; + indicates a definite, + + an intense reaction.} : ‘Symbiotic ” cultures, bouillon Renu One oF Pure cultures, bouillon clear. showing growth, the nitrite as medium. | | No. 25. No. 39. No. 67. No. 49.| No. 29. | No. 12.| No. 21. Nitrite ...... adr || shar |] apap || apap || apap || ab oe 0 OF) 0 0 ee be Nitrate ...... 0 0 0 (0) (0) 0 ++ ieee ++) ++ + | + Cultures No. 49 and No. 29 were strains of culture “d’’; cultures No. 12 and No. 21 were of different origin. { With cultures still containing nitrites, evaporation with ammonium chloride preceded the diphenylamine test for nitrates. 266 Nutrification with reference to the Purification of Sewage. Table [X.—Experiments upon the Absorption of Ammonium Chloride from Dilute Solution by Various Finely Divided Solids. Original solution of NH,Cl added. Liquid after standing, ammonia content per Ammonia in blank control in thousandths UL OSEGD joists. of a milligramme. No Solid tested ; weight taken. hone Volume content Upper Lower : of liquid per clear turbid i. ae foe in c.c. | 100,000 liquid. liquid. ques |) saute parts. 1 BaSO,, 1 gramme 100 1:00 1°02 0°91 0 0) 2 | Sand, 2 grammes ...... 100 5 ‘00 4°30 4-04 2 0 3 | Sand, 2 Pe eee 50 5 00 4-14 3°78 3 (0) 4 | Clinker, 2 grammes ... 50 1-00 0°82 0°89 2 0 5 | Sand, 2 grammes ...... 50 1-00 0°75 1-03 5 0 6 | Sand, 2 san eliiys abiis 50 1-00 0°83 1°05 3 (0) Th | ISEMICY cco cassedescnsescne-0 — 1-00 0°85 0:99 In Experiments 4—7, 10 c.c. N/1 KHO was added before distilling off the ammonia. 267 The Action of Anesthetics on Inving Tissues. Part 1—The Action on Isolated Nerve. By N. H. Aucock, M.D. (From the Physiological Laboratories of the University of London and St. Mary’s Hospital Medical School.) (Communicated by A. D. Waller, F.R.S. Received November 9,— Read December 14, 1905.) CONTENTS. PAGE Ti, UiatirroehCBKOr ..ocoodcoodaoadonnocongonhangadoeceodso0or BeoasedaCobangoD wugeoudanbasodupan60000 267 II. Experiments.—Series I. Anesthetic to Whole Nerve.................00ceeeeees 267 Ill. Series I1.—CHCl, to Parts of Nerve in varying Percentages ............... 269 IV. Series I1I.—Fixed Percentages, CHC], and Ether ................:scseeeeeeeees 271 V. Series [V.—Simultaneous Anesthetic “ Balance” Experiments ............ 273 VI. Series V.—Electrical Resistance of Nerve after CHCl,, and Control IX PELUME NUS prreccceeecee eee hce mee sos etesch once ser cleececenaeccesscaelsaceacle 275 VII. Series VI.—Electrical Resistance of Nerve after Ether ...............:000.0008 278 Wa lelibie Observations iivnssssachschenasscesse ce satins cecuesaecuasseasei odie avecevassueraaaneseens 280 Introduction. The action of anesthetics on isolated nerve has already been studied by Waller* as regards the effect on the negative variation in the sciatic of the frog, and by myself in mammalian nerves.t In the course of these researches it became evident that the anesthetics used (chloroform, ether, carbon dioxide) affected not only the negative variation, but also the injury current, and as this action has not been studied before, as far as I am aware, it seemed desirable to investigate the matter not only in nerve, but also in other tissues. ' The inquiry falls naturally under two heads: first, in how far the phenomena throw light on the processes of nerve action ; and, secondly, as regards the chemical and physical action of anesthetics on the animal protoplasm generally. These are obviously only parts of the same story, but for convenience I have considered the subject mainly under the first heading in this part, leaving the more general question to a future occasion. Experiments. If the sciatic nerve of a frog be taken and a fresh transverse section be made at the distal end, and this end be placed in contact with one non- * Waller, “Lectures on Animal Electricity,” 1897, and ‘Proc. Physiol. Soc.,’ November 13, 1897, ete. t ‘ Roy. Soc. Proe.,’ vol. 71, p. 264, and vol, 78, p. 166. VOL. LXXVII.—B. U 268 Dr. N. H. Alcock. [ Nov. 9, polarisable electrode, while another electrode rests on an uninjured longitudinal surface, the galvanometer will indicate a difference of potential which declines at a certain rate. This can be measured either by photo- graphing the movement of the galvanometer spot, or, more accurately, by balancing against a known potential, reading the potentiometer at convenient intervals of time, and plotting the figures on squared paper; this latter method eliminates any change in the resistance of the object. The curve obtained is concave to the abscisse, neglecting a variation which is occasionally found during the first five minutes of the experiment. If, when this curve has assumed a typical form, CHCl; vapour be applied to the nerve, a sudden drop is observed. If the vapour is weak, this is followed by a recovery when the CHCl; is removed, if the vapour is strong (12 per cent. or over) the drop is permanent (fig. 1). This may be due Li - 200 Fs 190 f Ee 180 - 170} (106) 160 Fie. 1.—Sciatic of Frog. CHCl, vapour at 12°5 per cent. between the vertical bars. Ordinates 1 mm. = 00001 volt. Abscissee 1 mm. = 15 seconds. 1905. | The Action of Anesthetics on Living Tissues. 269 to events occurring either at the cut end, or at the longitudinal surface or both, the experiments in Series IJ were undertaken to determine which of these factors predominated. Series LT. The nerve-chamber was divided by a transverse gas-tight partition of modelling wax, so that the aneesthetic could be applied either to the cut end A or to the longitudinal surface B (fig. 2). The figures obtained from the To sec. coil Y, d Y Fic. 2,—Nerve chamber. potentiometer readings have been plotted out to a scale on the ordinate of which one division = 0°0001 volt, and on the abscissa one division = 15 seconds. The point at which the anesthetic was applied was taken as the zero, and the relative heights plotted from this, the values are, therefore, relative and not absolute (fig. 3). The figures under the middle part of each curve indicate the percentage of CHCl; vapour, as measured by Waller’s densimetric method.* The vapour was contained in a bag of gold-beaters’ skin; the leakage from this was ascertained to be small during the time taken by an experiment of this kind. Here, as elsewhere in this paper, I have only considered experiments where the results are sufficiently concordant to admit of a consecutive series being employed without having to reject isolated experiments. Any unsuspected errors are, therefore, of a constant magnitude throughout a given series. It will be seen that— (1) CHCl, to cut end of nerve causes an increase in the injury current. (2) This increase is roughly proportional to the strength of the CHCl; vapour up to about 12 per cent., greater concentration then gives no further increase. (3) CHCl, to longitudinal surface causes a decrease of the injury current. * Waller and Geets, ‘ British Medical Journal, June 20, 1903. U 2 Exp.926 G90) +- EXp.929 pit \ao% fe) © Oo 6 iS Exp.927 (126%) fo) fe) “I Q © i fe) O (o)) Ov Q ce) © — Exp.930(17-9% Tes 5 etc. minutes. Fig. 3.—Series II. Relative increase and decrease of injury current by CHCl, vapour at different strengths. Ordinates 1 mm. = 0:0001 volt. Abscisse 1 mm. = 15 seconds. -OLO The Action of Anesthetics on Living Tissues. 271 (4) Recovery takes place when the vapour is weak (9 per cent. or less), not occurring when the strength is 12 per cent. or more. Series ITI. The method adopted in Series II, while suitable for percentages of CHCl from 1 to 10, is defective in many ways, and an apparatus was designed to deliver a constant high percentage of CHCls, saturated with water vapour, and at a constant temperature. This consisted of a foot bellows with a branched delivery tube, one limb of which passes to seven Fresenius flasks, the first three containing CHCl; the next water, the next two contain a mixture of CHCl; and water, and the last is empty. The apparatus delivers moist CHCl; vapour at a concentration of 13 to 17 per cent. (depending on the temperature of the first flask) with a variation in strength of about 0:5 per cent. and a temperature variation of about 04° C., and by estimating the strength of the vapour after it has passed through the nerve-chamber, and noting the temperature before and after each experiment, these residual errors can be allowed for. The five flasks on the other limb of the air supply contain water, and a small stream of air passing through them serves to ventilate the other part of the nerve-chamber, thus guarding against accidental escape of CHCl;. A similar apparatus is used for ether: in this case the percentage is about 45. Unless otherwise stated, all further experiments were made with this apparatus, with occasional modifications. In this series the strength of the CHCl; vapour was 12°5 (-0°5 per cent.), and was applied for 10 minutes. The details of the various experiments differed. The values are now absolute, and have been corrected for electrode eurrent. Each ordinate division = 0:002 volt; each abscissa division = 15 seconds. In Experiments 108 and 107 (nerves of Frog A) the CHCl; was applied first to the longitudinal surface and then to the cut end. In Experiment 109 (Frog B) the same course was followed, only without making a fresh section, the curve therefore begins close to the zero line. In these three experiments the curves fall at first, the effects being naturally opposite in sign to those in the next three experiments—Nos. 111, 110, and 117—where the cut end is first affected. No. 110 is practically a repetition of Series I, No. 111 ‘the same without a fresh section, and No. 117 shows the effect of ether applied for five minutes instead of ten. Considering the first half of each experiment— (1) CHCl (Experiment 111) and ether (Experiment 117) to the distal end of a nerve cause a rise in potential, approximately equal in amount. Volt CH Cl; CH CL, -020\—,—_ #4 I 234 5 minutes Fig. 4.—Series III. CHCl, of 12°5 per cent. and ether of 50 per cent. to either end of the nerve alternately. Ordinates 1 mm. = 0°0002 volt, abscisse 1 mm. = 15 seconds. Expts. 107 and 108, CHCl, to L. surface and then cut end. Expt. 110, cut end and then L. surface. Expts. 109 and 111, CHCl, to different ends without a previous section. Expt. 117, ethers 50 per cent. to cut end and then L. surface. The Action of Anesthetics on Inving Tissues. 273 (2) The effect of a previous section (110) is to produce a somewhat larger rise. (3) CHCl; to the proximal end produces a fall in potential (107, 108, 109). When there is no previous section (109) the experiment is obviously identical with Experiment 111; when there is a previous section, the total effect is less. In other words, if « = effect of section, and y effect of CHCl; in the first part of each experiment, Exps. 109 and 111 = y, LO hs a7, ne = 2+Y. These results are discussed subsequently in connection with the question as to whether z and y are due to the same mechanism in the nerve or not. Later experiments (Series V and VI) point to an affirmative answer ; as regards this present series, it appears that y is of the same order of magnitude as 2, whether CHCl; or ether is used, and that y has the same direction of current as w. A few experiments were made with alcohol vapour; this causes an effect of the same character as CHCl; and ether, but as it also causes an enormous increase in the resistance of the tissue, due to drying, the experiments were not proceeded with further. When, in the second part of each experiment, the anesthetic is applied to the other part of the nerve, the final result is similar to that of Series I, where the whole nerve is at once anesthetised. As can be seen, there is a small residual E.M.F., occasionally (in other experiments not quoted here); this may reach a value of 0:005 to 0:006 volt, the exact significance of this is still uncertain. Series IV. In this series, CHCl; and ether vapours of different strength were applied simultaneously to both sides of the nerve. The result could have been pre- dicted from an examination of Series I and II, but is a useful control, as certain sources of error are eliminated. Experiment. | Central End. Peripheral End. | E.M.F. initial. E.M.F. final. | | per cent. per cent. 931 51-0 (ether) 16 ‘1 (CHC1l,) +33 -— 1 932 | 16:1(CHCL) | _ = § —238 933 | 12-5 (CHCl) | 16-5 (CHCI,) +13 =e a 8:0 (CHCl) | : —184 (a) o34{ § | 43 (oHeL) jaa 5 (CHCl) | —16 ate (a) 113 | 54-B2°5 (ether) 12-3 (CHCI,) | —25 = 7 274 Dr. N. H. Alcock. [Nov. 9, If the small final values under 20 be disregarded, these experiments show that :-— (1) Ether vapour of 51 to 54 per cent. and CHCl3 of 16 to 12:3 per cent. are identical in their final effects. (2) CHCl; of 8 per cent. has a less effect. » CHCls vapour of about 12 per cent. has, therefore, a maximum action. (3) Ether acts more quickly than CHCl, (fig. 5). Fic. 5.—Series IV. Control Experiments. In order to exclude any possible source of current from electrodes or any unexpected physical cause, control experiments were made with threads moistened with (a) salt solution, (>) egg albumen, (c) egg yolk, and (d) fresh blood. These were treated in an exactly similar manner to the nerves in Series II. No electromotive phenomena followed the application of CHCl; vapour. So also nerves kept in salt solution until they gave no injury current showed no results with anesthetic vapours. HCl and NH,OH gave electromotive effects when applied either to threads or nerves, HCl with the same sign aS CHCl;, NH,OH with the reverse. The similarity is therefore only apparent and probably due to diffusion into the electrodes. 1905.] © The Action of Anesthetics on Inving Tissues. 275 Series V. The effect on the electrical resistance of the application of anesthetic vapours is of great importance in endeavouring to give an explanation of the facts recorded above, and so, although the question has already been studied by Waller,* I have performed a considerable number of experiments in order to eliminate as far as possible the numerous sources of error present. The method finally adopted was to place the nerve on non-polarisable electrodes in a chamber, pass in air saturated with water vapour, and take several successive readings until either they remained constant or else the rate of change was determined. Moist CHCl; vapour was then admitted for five minutes, a reading taken, the chamber washed out with moist air for four minutes, and the final value observed. A linen thread wet with M/10 NaCl solution rested on non-polarisable electrodes in the same chamber, and readings of its resistance were taken alternately with the nerve. A thermometer as close to the nerve and thread as possible gave an approximation to the temperature; as will be seen, the mean variation of this is about 0°4° C., and in view of the difficulty of securing an equal amount of moisture and an equality of salt content in the electrodes, threads and nerves, the temperature error may for the present purpose be disregarded. Two methods of determining the resistance were used :— (1) The ordinary Wheatstone bridge, using non-polarisable electrodes and a constant current from one Leclanché cell. (2) A slightly modified Kohlrausch apparatus, using the same electrodes and alternate currents. With the control threads the results were identical; both sets of readings show a diminution of resistance of about 2 per cent. after CHCls, in part due to temperature alterations, in part to other causes. The nerves on the other hand gave markedly different figures with the two methods. * Waller, ‘ Proc. Physiol. Soc.,’ November 12, 1898 [ Nov. 9, Dr. N. H. Alcock. 276 6S8: 0 ae 1-91 , 86: O @A.L9 NT 088: O } = peeryy, 196: 0 991 EYRCING OLL- 0 } promt | 096-0 ox Cet OAL NT 606: O prety, 20: T } Soe Delt OALO NT 8S8: 0 } peemT, | 066-0 2a de ‘mmodea "S0}0NT a Whe *IOHO 70 od esequooieg a nan tC OO (q) syuounttedxnp 00s‘0T 0028‘S2T OOMIZIT | estou fF jo uUvoy, } £0. LT { } 89. 9T } (Vy) syueutredxa 008‘FOT 002‘90T | OOT'90T | peetyy H Jo uve 0OL FTL } @. 91 { 008‘0FT } LOR { O0r‘0ST snetigenneeess a 00S‘TIT 008 FIT 00g‘STT ave STD 00406 \ LAI { OOF L01 | 1.91 { OO LTE Waeenaeetes ceases a 009‘SOT 000‘80T QOOOET [ets “"¥ OLT 002‘98 i { 00e8‘ZOr | { O0FrS6 eet wee eee eee ees sete q 009'86 8. LT OOT‘F6 6: 91 COMA prrerrosoomreme re soen Y €or OOL‘T2ZT } { 00L‘0ST \ OOL'ZFL ee) q 00z‘S0T all 008‘Z0T 20 { OORLZOT [erste ¥ gor “aT IOHO 5 “OHO : Ty (sumo) 109J8 ey suLmp aa "1IQHO e10F0q “queued xa, eoUBISISOyT ©) Q0UBISTSOY © | Q0UBISISOY ‘aOpLIG, 9U04s}VOT AA. "A Selleg ‘ORIOG 8 S01qg@—' Vy 277 SSUES. ’ L, . ion of Anesthetics on Inving The Act 1905.] (q) syuouttntodx oy 686. 0 ae 00'TZT 000'L2T OONSSL | eAtou g Jo Woy GP. AT 84: LT 80: AT (Vy) squourtod x5] 616: 0 002‘¢6 000‘L6 008‘L6 prot g Jo Uno @ATO NT £96: 0 F j { 000'88T i { 000‘66T \ an { 000‘061 pret | 96-0 } Cael che 000‘L01 ren 000‘0TL PEM 000'TIT | ‘ ¢ G OAT NT £16: 0 \ : j { | 000°ZAT \ . { 000941 } { 000'9LT proryL | 186-0 aL8l oye 00024 Seat 000‘08 En 000'18 OALO NT 186-0 \ { 000°6L \ ; { 000°28 \ : { 00008 prong, | SFO. T Oyek O-8T 1} 0006 et 000'06 wad 000'88 AMO NT £20: T : { 000°48 } 3 { 000'06 \ { 000°¢8 eee ao pry | 96-0 i 0: 91 $- LT 000°82 8.21 GG0'UR 8: OL OOOTe tot serene Bog BALIN | 980. T Es o000'¢s \ { 0n0‘88 | { 000‘28 vevueeeer Ceeeen q prot, 916-0 j as { 000‘82T Us 000¢2T Dey COCO et avantr “modva : “f ; “f ’ *(sutyo) | “ny /t & en ee auutp ano erored t | ar ) -quourtodx 80}0N w/a TOHO 7° | : 04yn : au : TOHO ot0Joq quoutptod xy esuqyueIEg | oh BOUNASISOY | ow OOULISISOY, oh ODUBISTSOIT Yosnelpyoy “arog 8.so1qg—gq ‘pe ULpUOI— SOLIS 278 Dr. N. H. Alcock. [Nov. 9, These figures show that— (1) Nerves tested by Wheatstone’s bridge show an apparent diminution of resistance after CHCl; of about 15 per cent. (2) Nerves tested by the Kohlrausch method show no alteration of resistance. The difference between the two methods is that any polarisation in the nerve appears by the first as an added resistance. The figures in the first column of A therefore represent resistance and polarisation, and the conclusions follow— A. CHCl; diminishes the polarisation of nerve ; B. CHCl, does not alter the real resistance of nerve within the limit of error of the method. If the diminution in the control experiments is taken as a correction to be applied to the nerves, the resistance of these after CHCl; is increased by 1 per cent., at present it is doubtful if this correction is to be applied. -~- [Note added December 12.—A fresh series of experiments (Nos. 211 to 214) gave the final ratio for nerves slightly Jess than the control threads. The probable limit of error of the resistance experiments in the text is about +2 per cent., and as this lies in the object examined rather than in the measurements, greater accuracy seems at present unattainable.] Series VI. Similar experiments were made with ether, taking readings by both methods. The control threads now show a diminution of about 5 per cent. and an increase during etherisation. Except in this latter particular the nerves gave results almost identical with CHCl. (a) S F66: 0 ¥- ST 000s, | &.8T 00g FIT Z- ST 009°¢), | seAtou F Jo inn y a 676-0 F- ST 000°62 g. 8T 008‘ LOT 3: 81 00s*s8 = | PReTTY F Jo ULL OALO NT _— 696: 0 | : { 000‘eL, { 000‘9TT { 00022 sereeeeneee of proxy, = T96: 0 €- 8T 000°66 g- 81 000°Z8T &- 81 ooo'eoT te ¥ OTZ : OAL NT — 86-0 } { 00069 { 000°20T { 000°ZL ereeeennee gy & preryy, == 6cg. 0 6: 8T 00089 &- 81 00°88 8: LT 00099 ty BOR > ia) ¢ G a ¢ see eee eeeees az 9AON = 000- T } ; { 000‘¢8 : { 000‘62T : 00068 a ES promt, — 126: 0 ae81 000'e8 oe a ooo snr US stoner. [pecans 7 rae | 2 a4.t0 NI ae &L0- T } { 000'82 | { 000°60T Oo) — Re i -$ prea, = 096-0 9. 81 000°TL 8-8T 1 go0's6 ¥- 81 O00, ee LOR S ‘xoidde yueo rod 0g 3e OIA ‘Yyosneiyyoy ‘“peoryy, pure onerog sSo1g—gq ~ S 5 998: 0 = O16‘Z2T = 086°S9T OSTeFT .o aS aaee Se SS SSS SS ~ 0. 0S 698: 0 G. ST OOF CFT = 000‘86T L- 81 QOGIES Ia i ams nnaumes OOD — 8. Lb Zo8- 0 8: 8T 008‘TFT = 000‘06T 0: 61 O08;69 een |eeereenemmee OD 3B 0. TS IT6: 0 8- LT OOLISy = | = 009‘90T 3: ST OOP98) ea retrace eeeOn 8 0. ¢¢ 928- 0 6: LT o0s’6z— | OT 008‘69T 6- LT QO9'9ST rE TOT aTOTMO8TO ‘Sor hoses a SU QALOU 94} EpIssuo[e Md Oe Le See Ve S qgou pues ‘reqje ey <= | Jo amodeA oy ur uoye} L¥6- 0 = i CHG = 009°81T = OLL'66 Ss SBM oINgriedure, 94 a a ne Sane pa as Ss pus ‘{yJeatsseoons porns FBS 696. 0 G. 8T 006‘00T = 000‘ L821 1. 8T oogsot ee JOT = “BOUL JIM SpvaIqy PUL 0- 0S £68: 0 P. ST OOFOTT a 000‘EFT 9. ST OOF'eZr jc GOT SOATOU OY} SeLIos SsIq} UT 0- OF 066: 0 0-8I | o00TZ = 000°98 &- ST OOL‘TL teraeseeseseess DOT d | ° | ~~ I S ‘OOPLIG ooys}voy A “LOYY ‘spvorlyy, JoryWOQO— vy paasit slay — , : 4 5 Wy = suo S tee aes mate i pelgeene : RAED ; ‘Jeyqe aT0yo “quotartedx = sh asByue010g ale ok eoUB4SISey od eoURySIse yy ok pee 4 4 ; a =! “LAD SeHeS 280 Dr. N. H. Alcock. [Nov. 9, These figures show— (1) Ether diminishes the polarisation in nerve. (2) Ether does not alter the real resistance of nerve (with the same reserve as to the control experiments). Comparing the result of the two anesthetics, and taking the mean figures :— Mean resistance | Mean resistance R./R before. after. a/ Ry Nerves CHC wrrcrerecters 121,400 103,500 0 *854 Wheater se Veber so ae ee: 143,180 122,910 0 865 Care nee Nerve, CHCl, ............ 122,600 121,200 0-989 : De eal 75,500 75,000 | 0-994 | Kohlransch. The final effects of CHCl; and ether are identical, within the limits of error of the experiments. Observations. The full discussion of the problems raised by the facts here recorded must be deferred until the results are considered of experiments on other tissues and on the question as to the actual effect of the action of CHCl; and ether on proteid solutions.* Taken as they stand, the present experiments show :— (1) That chloroform and ether (and probably alcohol) produce an electro- motive effect when acting on a frog’s nerve, which has a maximum value of about 0:030 volt, and the same sign as the current of injury. (2) That CHCl; and ether produce no alteration of the resistance of the nerve (within the limits of error), but diminish the polarisation. Two inferences present themselves— A. That the electromotive effects are due to the same cause that produces the injury current. B. That as the resistance is not diminished no additional ions are formed. The discussion on the correctness or otherwise of these inferences is post- poned for the reasons given. I have again great pleasure in acknowledging the kindness of Dr. Waller, both for the permission to work in the laboratory of the University of London, and for advice in the conduct of the experiments. Also to express my appreciation of the assistance of Dr. B. J. Collingwood, in conjunction with whom many of the experiments were carried out. Mr. Shapiro has also given me much help in the later stages of the research. * See Moore and Roaf, ‘ Roy. Soc. Proc., and also Waller, doc. c7t. | Ho OHO Ho “IOHO | H° "OHO #9 "IOHO | #9 “IOHO Eas PZ — FOS — 66 + 0 BAT = oh VL + a) $9 cog | O9T + 16 ed A ecg G08 6ST + 96 eL oc 908 8g o6 ah Gg 908 8st ¥6 BL CG 108 Sgt 26 TL | LS 808 eg 88 OL = ; | Se 808 OST 98 89 Si dquo woryvordde ‘suru g | BSB 108 SPI 8-91 = al 08 99 > HO LOI 4 908 ost GL 89 2 993-— 6h FOS OFT ‘aip g = °° 49 8¢ N 19% | LPS B08 aa 09 ag 19% | 8S 962 rea 9F tH = eter. 4 886 062 ogt og + ag Ss 9-0 = **P* og | L@e 082 SIL M3 so + -S BLE ed 898 COT oz — 8 + N OLB | 681 PSS 6 os — 0g = 2 0s@ | O9T e8z 84 + 8 — oF — S 008 — 0zI— 81Z— og + PIL — 08 — zx OZI— 06 — 06T— oe + PI oZI— Ss “pus [esp OF eq} =| = —*pue Teystp *1OHO “pue yno 09 *IDHO ‘pue ‘xoad 09 "IQ HO "Fms “SUT 04 "OHO ‘Fans “SUT 09 "1OHO Y 2 J a: es 8 191— = cc ot Lig 891 12 te x aa LZ SLT gers 691 SPL ral o% 9LT a TAI ial = aI GS OLT 8% BLT PT rae 9% GAT &% PLT 8-041 = ob OPT S pue yn ou TT pus ynd ou 1Z GLI &% SLT SPL Ss IL 1 GAT pue yo ou Bz LL 6FI 38 Or 9% cL — 1% 1.3 = **2* 6L1 TST NX ooot 6 is 9 = OOoOoT si — ooot Te = est — Vallis S 02= ig — cs — Ceo re Oe gig oe iee— *8I— 991— i N ‘S— juoLmMo eporlgo9[y | “S— jyuatmmMod epo1qgoe [A e+ = 3UILMI epor1g09[H |} “S— quetmo aporqoo TF ‘Q— JUueIIN) sporqZoI[ "E@— JUOIMMO Vpo1pDOTA “LIT queued x ay ‘TIT queurnedx ay ‘OIL guowmniodxa "601 quourttodxgy ‘SOT quoutttedxay "LOT yuounrsodxgy re oS “TOI “10HO lor) — ‘S[CATOJUL “O08 YE JV 9AIOU OY} WOIJ QUOIIND oy} SuToURTeq ‘1ojoTIOI}UI30d 9Y4 Jo esoy} OV UBATIS SaInsy OYJ, “ARE SO8T19G ul syuotted xq jo S[000901q *XIGNHdd V . OOOO ee [Nov. 9, Dr. N. H. Alcock. 282 ATuo uowyeordde ‘sur ¢ HO 1949, ZL — 8 II 91 8% LE 09 101 ] 06 — *pua xord 03 10q4qy 0 *10HO 6 — 48 60T O&T SEN 0ez— ‘puo ‘xoad 03 "OHO 90 "I10HO rE og ge 98 8g 1 PP Le #S 29 — oi, = 8 — oor— eZI— ost — O8T— 0% — Goo — L-9L = ob ¢8¢— ‘Jans “Bu0] 04 *IOHO H° *IOHO L@ 8Z 82 83 82 82 62 62 0g 62 0g og — 62 — TS — Mt — 9 OT re 8g apap ap || “pue [BIsIp 09 [OHO ‘B° *1OHO 8..9T = of AT — Je = AT Zt 41 GP 99 ‘pues yno 03 *[OHO + 06 + + + a a ee ee 8° "I1OHO 7 Ve &Z &% &Z T-o41 — os &3 &Z &@ &Z & &% ‘AIP 8.6 = 299% og ¥S 7 — fab = 02 — OL — ets 02 + OF + *pue qno 09 *1QHO 093 — 196 196 696 696 £96 £96 £96 9S 996 — ‘8— JUeImMd oposqooT OT “LIT quountaedxcy 9FS— 9G 9G LVS 876 8S 606 Oss 1S@ €sg— *g— queTmMo opo.oea[y ‘TIL queunsedxay 866 — 866 666 666 666 008 00& 00€ T0€ €0& — ‘p+ = JuetmMo opoaqoo[q ‘OIT Juoutsedxay 8S + 8ST 8ST LST LST LST IST 8ST O9T + "8— quero oporo9 [A “60L quoursedxny 86 + 86 86 86 86 86 86 86 86 6..9T = of 86 + 68 + 68 *9— JUeIIND epomqo.9[ FT "S01 queued *E— qUeIMD epoaqooy ay "LOL Juowrtedxgp “TOUT ‘panurjuor—] J] Sol1eg Ul syuewIIEdxG Jo sjooojo1g ““10HO "XIGNHdd Vy 283 Tissues. wing L wcs On ion of Anestheti The Act 905.] ils = queLIno apomoe[qy WOOWODD 9— quero sporqoo[oT &— — AODDIr~@ are AL = Wi = JUSOLIMd opo1po9e[A VG &Z 96 8g — ‘gAcOU O[OTM 03 *[Q HO 9.g == 200% ‘6— = quero aporqoa[ €& €& && — es = *aATIM CTO 04 *[QHO f— = quarto aposoopy LI LT 9T Ny = ‘gAdou O[OTM 07 *IOHO G— = UII 9po.1q09/ FT GS (4G 02 — ‘dALOU BTOTM 09 "OHO B. VOL. LXXVII. 284 Further Work on the Development of the Hepatomonas of Kala- Azar and Cachexial Fever from Leishman-Donovan Bodies. By Lreonarp Rogers, M.D., F.R.C.P., F.R.C.S., B.S., Indian Medical Service. (Communicated by Sir Michael Foster, K.C.B., F.R.S. Received October 16,— Read December 14, 1905.) [Puate 7.] In 1903 Lieutenant-Colonel Leishman, R.A.M.C., described certain bodies in the spleen of a fatal case of chronic fever in a soldier invalided home from near Calcutta, which he considered to be degenerate trypanosomes, on account of their resemblance to the breaking-up dead trypanosomes found in the spleens of rats 48 hours after death, and he therefore suggested that trypanosomes might be present during life in this class of fever. Major Donovan, I.M.S.,(2) of. Madras, however, very shortly after showed that Leishman’s surmise was not correct, as he found similar bodies to those described by Leishman in fresh blood obtained by puncturing the spleen during life, but no trace of trypanosomes; and after examining Donovan’s specimens, M. Laveran (3) pronounced the parasite to be a Piroplasma, and suggested the name Piroplasma Donovani for them. Donovan (4) also claims to have found the parasites in the red corpuscles of the peripheral blood, but his coloured illustrations of them very closely resemble ring-parasites of malaria, and have only one chromatine body, and his statements in this respect have not been confirmed by any other observer. Ross (5) suggested that the parasite probably belonged to a new genus, and proposed to call it Leishmania Donovani, and Manson and Nuttall also favoured the view that it is a distinct genus. In the following year, 1904, on my return to India from leave, I commenced an investigation of the subject, with a view to finding further stages of the life history of the parasite which might throw light on its true nature and classification. In the meantime Lieutenant Christophers, I.M.S., had been placed on special duty by the Government of India to investigate the subject, and after making a careful study of the parasite in different tissues of the body, he suggested that they might be spores of a microspordium. (6) My first endeavour was to find some method of keeping the parasite alive outside the human body, and after a number of trials success was obtained by preserving spleen blood containing the parasites under sterile conditions in a cold incubator, preferably at about 22°C. For this purpose the fresh blood obtained by spleen puncture during life was placed in small sterile Development of Hepatomonas of Kala-Azar, ete. 285 test-tubes containing a few drops of 2 to 5 per cent. citrate of soda in normal salt solution, in order to prevent clotting. Under these conditions not only did the parasites remain alive for many days, but they also multiplied very rapidly, became much enlarged, and after about three days some of them developed into elongated flagellated bodies, which I took to be a stage in the development of a trypanosome, although no undulating membrane was yet present. This discovery was announced in the ‘ Lancet ’ of July 23rd, 1904, a few uncoloured illustrations were published in September, (7) and a fuller paper tracing the stages of the development day by day, with a coloured plate, in November (8) of the same year. Confirma- tion of my discovery was first furnished by my assistant, Dr. G. C. Chatterjee,(9) working in my own laboratory, and next independently by Lieutenant Christophers, (10) working in Madras. Thirdly, Captain Statham (11) and Lieutenant-Colonel Leishman also obtained the development, and published their account of it, March, 1905. During the past year I have made a large number of experiments on the conditions affecting the development of the parasites outside the human body, with a view to obtaining a clue to the natural mode of infection, and early in the present year I published a summary of the results obtained. The most important conclusions arrived at were, that sterility is essential to the continued development, and that flagellation takes place much more uniformly and regularly if the citrated spleen blood is faintly acidified with citric acid: facts which strongly point to the stomach of some blood sucking insect as the natural place of development of the parasite outside the body, and I gave some clinical reasons for considering the common bed bug to be the most likely conveyer of the disease. So much more abundant develop- ment of flagellated stages has recently been obtained by the use of acidified blood medium, that I have been able to make a more satisfactory study of the exact mode of development, and to come to a definite conclusion regarding the ultimate stage it reaches, and therefore propose to describe and illustrate these later stages more fully than in my previous papers, and to briefly discuss the bearing of the conditions affecting the development of the flagellated stage of the parasite on the probable mode of infection of the clisease. Stages of Development of the Parasites observed in Acidified Citrated Blood. In the first place, the development in acidified blood is much more uniform than that obtained by the previous method, so that instead of finding all stages present after three or four days, with a great preponderance x 2 286 Dr. L. Rogers. Development of Hepatomonas of [Oct. 16, of the smaller oval forms, and but few flagellated ones as in alkaline blood, in the acid medium the great majority of the parasites will be found in nearly the same stage on any given day, and nearly all become flagellated after a few days. The sequence of events during the first two days is the same as I have already described, (8) and they are well shown in the first two lines of the accompanying plate—all the figures in which have been drawn to the uniform scale of 1500 diameters magnification, with the aid of a cainera lucida. Line I shows the parasites seen in a film of spleen blood made at the time it was obtained, and consequently before any development had taken place. After incubation for two days at 22° C. the forms shown in Line II were present, figs. 1 and 2 showing considerable enlargement, especially of the macronucleus and protoplasm of the body. Figs. 3 and 4 also show the earliest appearance of the eosin-staining body, which is repre- sented as a clear space in the drawings, but is of a rosy-pink colour with Romanosky’s stain, and quite distinct from the vacuoles, the latter being indicated by the more lightly shaded portions of the protoplasm. It will also be noted that from the first the micronucleus, or blepharoplast, is closely attached to the eosin body (called by Leishman “ flagellar body ”). Further, on the second day in this acid culture a few of the early flagellated forms shown in figs. 7 and 8 of Line II were also seen, although they do not usually appear in alkaline cultures until at least the third day, while just antecedent to this stage are the forms shown in figs. 5 and 6 of the same line, illustrating commencing elongation and division by fission, and it will be observed that in these the eosin body is passing up to the anterior end of the organism from which the flagellum arises, and is carrying the micronucleus with it. In my earlier description I suggested that the double elongated forms shown in fig. 6 of this line might possibly represent a form of conjugation preparatory to the development of the flagellated stage of the organism, but further study of a much larger amount of material has convinced me that they are only fission forms, as I have been unable to make out any reduction in the number of chromosomes in the macronucleus during the process. The Mode of Division of the Flagellated Forms and the Formation of Rosettes. In Line III of the plate are represented the different stages of division of newly formed flagellated bodies. Figs. 1 and 2 show that the micronucleus and flagellum first divide, just as in trypanosomes, and next the macro- nucleus divides in turn, and a clear line appears in the length of the organism, indicating commencing division of the protoplasm of the body, as 1905.| Kala-Azar, etc., from Leishman-Donovan Bodies. 287 shown in figs. 5 to 5 of the same line; while in fig. 6 the division of the body has just been completed, and in fig. 7 the micronuclei and flagella of a still adherent pair are dividing over again, thus showing how rapidly the multiplication was taking place in this culture, for such forms were not uncommonly seen in it. In my earlier cultures the flagellated pairs were nearly always found in pairs only, although rarely three or four might be seen side by side. In the much more abundant development of flagellates in the acid culture medium, however, considerably larger masses, forming beautiful rosettes, with the flagella crossing each other in the centre, were seen in large numbers, and it is easy to understand how they may be formed by the rapid multiplication just described. Thus fig. 6 of Line IV shows a small group of flagellates which is remarkable for including nearly all the stages of division in a single clump, while fig. 9 shows the commencement of the formation of a rosette by the rapidly dividing flagellates pushing each other round to form a semi- circular mass, and in fig. 12 is shown a small, but complete, rosette, several of the forms in which are undergoing further subdivision. In this stage the contents of the eosin bodies frequently becomes protruded, as I have previously noted, and it accumulates round the flagella, helping to bind the forms together into the rosette shape. Next, the individual organisms elongate, and at the same time become narrower, and the rosette then commences to break up, in consequence of the increasing motility of the flagella, and some now separate from the mass in pairs or single forms as indicated in fig. 11, and in this manner the free swimming forms shown in Lines IV and V of the plate are produced. In fresh specimens these are very active, the single ones in particular threading their way rapidly among the red corpuscles, and on reaching an open space, dart about in such a manner as to leave no doubt in the mind of the observer that the object of this remarkable development and extraordinary increase in size is to endow the motionless human stage of the organism with the power of loco- motion required in some period of its extra-corporeal existence. The Nature of the Fully Developed Flagellate Form of the Organism. When I first obtained the development of the flagellated stage, I thought them to be young trypanosomes which had not yet formed an undulating membrane. In support of this possibility, the recent observations of Novy and MacNeil (12) on the culture of trypanosomes of birds on blood agar are of great interest, for they obtained forms, separate and in rosettes, most closely resembling those shown in the plate accompanying the present paper, 288 Dr. L. Rogers. Development of Hepatomonas of [Oct. 16, both in the absence of all trace of undulating membrane and in the position of the micronucleus or blepharoplast at the anterior flagellated end of the organism, although in addition they obtained forms showing the development of the membrane by the passage of the blepharoplast back towards, and then past, the macronucleus, until it arrived near the posterior end of the organism, and a typical trypanosome resulted. When further experience of my culture failed to reveal any forms with a complete or even partial undulating membrane, the question arose whether it was not an organism distinct from the trypanosomes, although closely related to it, such as a hepatomonas in which no undulating membrane is present. In my last paper I left this an open question, while stating that nothing had yet been found which might not be an incompletely developed trypanosome; a view which has also been adopted by both Christophers (10) and Leishman (11), The more abundant and uniform development of flagellates in the acidified medium have enabled me to study closely innumerable apparently com- pletely developed long free forms, in a stage in which they show extremely active movement in fresh specimens: but still no trace of an undulating membrane, or even a tendency for the micronucleus to pass away from the anterior end of the organism towards the macronucleus has ever been observed, although seen by Novy and MacNeil in their cultures of bird trypanosomes. I therefore conclude that the organism I have been able to develop belongs to the order Hepatomonas and not to the trypanosomes, and I propose to name it the Hepatomonas of Kala-azar. At the ‘same time ii. prefer to limit the term kala-azar to the epidemic-spreading form of the disease as seen in Assam, and to retain the term “cachexial fever ” for the less fatal sporadic affection, if only for the sake of avoiding the unnecessary eruelty of having to tell sufferers from the milder disease that they are suffering from the greatly dreaded kala-azar. Degenerate Fors. I have already pointed out that the absence of bacteria is necessary for the continued development of the flagellated stage of the organism, and that cocel especially are inimical to its growth. In one of my most active recent cultures staphylococci gained access to the tube on the seventh day of the culture during its repeated examination, and the degenerating changes resulting were readily followed. On the following day fresh specimens showed that all motion of the flagella had ceased, although on staining many of the organisms showed little or no change. Others, however, were granular and stained more lightly, while some were becoming shorter and more oval or 1905.| Kala-Azar, etc., from Leishman-Donovan Bodies. 289 pear-shaped, and their flagella shorter, as in figs. 2 to 6 of Line VI, thus showing a tendency to reversion towards the undeveloped spleen stage of the parasite, only all stages of the degenerative process were present at the same time, and many of the shrunken badly staining forms were disintegrating. During this process the flagella were often shed, and with it the micro- nucleus came away, although a narrow non-staining space was still visible between the two, as shown in fig. 10, clearly proving an organic connection between the flagellum and the micronucleus or blepharoplast. Within three days all the rosettes of flagellates had broken up into granular masses and their identity completely lost. The degenerative changes in this hepatomonas are therefore very similar to those which have been described in the case of trypanosomes. The Lelationship of Leucocytes to the Parasites in Cultures. Although it is doubtful as yet whether the Leishman-Donovan bodies can be found in the peripheral blood either free or in the red corpuscles, yet both Donovan (+) and Christophers(13) have found this stage of the parasite within leucocytes in the circulating blood during high fever, the latter having twice found a number of them, nearly all within polymorphonuclears, during a differential count of 500 leucocytes, which would mean an enormous number within the peripheral circulation at one time, and amply sufhcient to infect a blood-sucking insect if such proved a suitable host. It is therefore of interest to determine if the parasites can develop in acid cultures within leucocytes. Figs. 7 and 8 illustrate conditions bearing on this point, the former representing a polymorphonuclear on the second day of the culture, which contains typical parasites, although they are somewhat less developed than those shown beneath it from the same slide: the latter shows another degenerating leucocyte from the same culture on the following day, in which some of the parasites are clearly much enlarged and developing typically, if somewhat more slowly, than those show outside the corpuscle, while others are degenerating and staining feebly. It appears then that development may proceed within leucocytes, while Christophers is also of opinion that it oceurs within macrophages in cultures, so biting insects might be infected by the leucocytes containing the undeveloped parasites which have been found in the peripheral blood. The way in which the polymorphonuclears especially take up the parasites in the peripheral blood is also of great interest in connection with the extreme decrease in these corpuscles, for I have shown that they are commonly decreased a tenth of the normal number, while in the latter stages of the disease, in children especially, they may fall to only from one- 290 Dr. L. Rogers. Development of Hepatomonas of [Oct. 16, twentieth to one-sixtieth, thus readily accounting for the frequency of terminal infection by such diseases as dysentery, cancrum oris, pneumonia and phthisis, owing to loss of phagocytic power, while I have also found the opsonic index reduced against the staphylococcus pyogenes aureus, which is frequently present in the spleen in cases of cancrum oris. The Bearing of the Flagellation of the Parasite in Sterile Acid Medium on the Probable Mode of Infection. The two factors which I have found most essential to a uniform develop- ment and very rapid multiplication of the flagellated forms are sterility and a slightly acid, or, at least, a neutral medium. I have also tried blood agar after Novy’s method, only using human blood im its preparation, but failed to obtain either sub-cultures of already developed flagellates or of the spleen parasites, while only very scanty development was obtained when several drops of spleen blood, with very numerous parasites, were added to a previously acidified blood agar tube, and then only in the added blood as by the ordinary method. Now the only condition under which the Leishman-Donovan bodies would be likely to meet with a sterile acid medium on their escape from the human body would be in the stomach of some blood-sucking insect, of which the common bed bug, or possibly mosquitos, are the most likely hosts, for clinical reasons I have elsewhere pointed out, while I have found that after sucking blood it becomes acidified in gastro- ntestinal tract of bugs, and is also frequently sterile. I have not yet succeeded in inducing these insects to suck infected spleen blood placed in capsules of various kinds, but, on the other hand, I have mixed the contents of their stomachs after feeding on human blood (which was proved to be free from anything resembling any stage of the Hepatomonas of kala-azar) with about an equal quantity of spleen blood containing the parasites, and, after incubating in capillary tubes at 22° C., have been able to watch the development of the parasites day by day up to the flagellated stage under these conditions in those which remained sterile, but not when any bacteria were present. It is therefore clear that the conditions met with in the stomachs of bugs—and possibly also of mosquitos—are not inimical to the development of the parasite into the flagellated stage, provided the temperature conditions are suitable. The more difficult question whether opportunities for infection of such insects occur sufficiently frequently to account for the incidence of the disease remains to be considered. In the first place, it is conceivable that bugs especially might become infected from skin lesions containing the parasites, for these may occur on parts of the body little exposed to the bites of 1905.| Kala-Azar, etc., from Leishman-Donovan Bodies. 291 mosquitos, but, in my experience such skin affections are too rare to alone account for the frequency of infection. Further consideration will, I think, show that the difficulty in finding these minute parasites in the peripheral blood does not necessarily exclude the possibility of their occurring there in sufficient numbers to infect insects, especially during high fever, when they have been found in circulating leucocytes. In the first place, it has been shown, by Christophers especially, that the organisms multiply in the endothelial cells lining the blood-vessels of the internal organs, such as the spleen, liver, and bone-marrow, and when numerous, in films obtained by spleen puncture, they are frequently seen in groups in fragments of these cells, which during life must frequently rupture and set them free in the circulation, as is also proved by the same observer having found them in the blood of some of the large veins. It is further of interest to note that the endothelial cells of these very same organs are the principal sites of the deposits of malarial parasites in the internal organs, while I have also several times found Leishman bodies in the brain (where malarial parasites also occur), so that it is clear that they must frequently enter the circulating blood in considerable numbers. Secondly, the human stage of the parasite is so small that it would be ‘scarcely easier to find in the blood by micro- scopical examination alone that typhoid bacilli in that disease, although the latter can be readily obtained by cultural methods. The great difficulty of finding the human stage of the Hepatomonas of kala-azar in the blood, even if present in sufficient numbers to infect suitable insects, is well shown by Novy and MacNeil’s(12) experience of searching for trypanosomes in birds; for while they only succeeded in detecting this large actively moving parasite by microscopical examination of thick blood films in 8 per cent., nevertheless they cultivated the parasite on their blood-agar medium in 50 per cent. of the same series. Moreover, even when they found them by their movement in thick fresh films, yet in the same birds they frequently failed to detect them in stained specimens. How much more difficult would it be to demonstrate the minute motionless Leishman bodies, which can be only seen in thin stained films, even if they were present in relatively large numbers in the peripheral blood ? Thirdly, the extremely rapid multiplication of the flagellated forms in some of my recent cultures would appear to indicate that, in the presumably still more favourable natural conditions of the extra-corporeal stage of the parasite, a very small number of the human organisms would multiply to such an extent as to constitute a powerful infective agency. The only reasonable alternative to the hypothesis just set forth is the suggestion of Manson, Christophers, and others, that the organism may escape 292 Dr. L. Rogers. Development of Hepatomonas of {Oct. 16, 8 yi p from the body by means of ulcers sometimes found in the intestines, the granulation tissue ~f which contains the parasites, and they may thus reach water. Apart trom the great rarity of such infected intestinal lesions in my very extensive post-mortem experience of this disease in Assam and Calcutta, the fact that sterility is essential for the continued development of the flagellated stage of the organism appears.to me to make this mode of infection an exceedingly improbable one. Moreover, I have been unable to obtain any development of the organism in even sterile water kept at the most favourable temperature, while even in sterile acidified water similar negative results have recently been obtained. Relationship of the Optimum Temperature for the Development of the Flagellates to the Seasonal Incidence of Kala-Azar and Cachexial Fever. If the conditions I have found necessary for the development of the flagellate stage of the Hepatomonas of kala-azar afford any indication of the natural conditions under which it occurs, then the striking fact that the relatively low temperature of about 22° C., or 72° F., is essential to the process, would indicate that infection is only likely to take place in India during the colder part of the year. Owing to the fever in this disease lasting for many months, or even several years, with long intervals of little or no rise of temperature, while cases not infrequently begin very insidiously, patients presenting themselves with marked, but often unsuspected, enlarge- ment of the spleen, and a history of only a few days’ fever; it appears probable that the incubation period may be a long one, and the onset very insidious and indefinite. Nevertheless a clear history can often be obtained, and an analysis of the notes of a number of cases showed five times as many in which the symptoms first commenced in the six months from November to April as in the remaining six hot months of the year, so that the cold weather months, together with the very commencement of the’ hot weather, to allow for the probable incubation period, show a very marked preponderance of the infection. Moreover, Dr. Dodds Price, of Assam, informs me, as a result of his unique experience of kala-azar, extending over 15 years, that every case he has seen in Europeans began in the cold season, and that among his hundreds of native eases .he has noticed the same marked tendency for definite symptoms of the disease to first show themselves at that time of the year. The practical importance of this point in relation to the prevention of the disease is evident, while the close agreement of its seasonal incidence with the deductions from my experimental data is of considerable interest. It is also worthy of note that this disease is most prevalent in just those parts of India where the temperature conditions for Roy. Soc. Proc., B. vol. 77, Plate 7. 1905.| Kala-Azar, ete., from Leishman-Donovan Bodies. 293 several months of the cold season most closely correspond with that which I have found to be most favourable to the development of the flagellated stage of the Hepatomonas of kala-azar, namely Assam, Bengal, and Madras. On the other hand, the disease is much rarer, or has not yet been proved to originate, in those parts of India where the winter season presents a greater degree of cold, and the more favourable spring and autumn are very short. Much work will be necessary to test the truth or otherwise of the above hypothesis, but knowledge should mean power to prevent the most terrible of all tropical diseases in its combined very high mortality and slow death by inches, and as the most favourable cold weather working season is approaching, it appears to be advisable to put these observations on record for the benefit of other workers in this very important field of tropical medicine. REFERENCES. 1. Leishman, “On the Possibility of the Occurrence of Trypanosomiasis in India,” ‘Brit. Med. Journ., vol. 1, 1903, p. 1252. . ‘ Brit. Med. Journ.,’ vol. 2, 1903, p. 79. . ‘ Bull. de Acad. de Med.,’ vol. 1, p. 288. . “haneet,’ vol. 2, 1903, p. 44. . ‘Brit. Med. Journ., vol. 2, 1903, p. 1401. . ‘Scientific Memoirs, Government of India,’ No. 18, New Series. . * Brit. Med. Journ.,’ September, 1904. . ‘Quart. Journ. of Micro. Sci.,’ November, 1904. . ‘Lancet,’ vol. 1, 1905, p. 16. 10. ‘Scientific Memoirs, Government of India,’ No. 15, New Series. 11. ‘Journ. of Roy. Army Med. Corps,’ March, 1905. 12, * Journal of Infectious Diseases,’ vol. 2, 1905, p. 256. 13. ‘Scientific Memoirs, Government of India, No. 11, New Series. to w IOS oO co oO DESCRIPTION OF PLATE, Magnification of all the figures 1500 diameters. Iie. 1.—Undeveloped Leishman-Donovyan bodies from spleen puncture film. “4 Il.—Early stages of development, from two days’ culture in acidified citrated blood ; 1 and 2, body and macronucleus enlarged ; 3 and 4, first appearance of eosin body; 5 and 6, elongation and subdivision ; 7 and 8, first appearance of flagellin. » II1,Stages of division of the early flagellated forms. » LV.—Double long swimming forms. “r V.—-Fully developed long, free, active single cells. » Wi.—Degenerate forms. . WIl—Undeveloped forms in a white corpuscle. » VIII-—Early stages of development in a degenerating white corpuscle. , IX.—Stage in the formation of rosette. » X.—Separated flagella with micronuclei attached. » XI.—Rosette breaking up into free forms. , XNL£I.—Small complete rosette. 294 The Factors which Determine the Production of Intraocular Fluid. By E. E. HENDERSON and E. H. Srarwine, F.R.S. (From the Physiological Laboratory, University College.) (Received November 23, 1905.—Read January 18, 1906.) In spite of the very numerous researches which have been made during the last half century on the seat and mechanism of production of intraocular fluid, ophthalmologists and physiologists are still far from an agreement on the subject, and a review of the literature reveals many discrepancies in the experimental evidence which it is impossible to clear away without a re-examination of the whole subject. The following paper contains the results of experiments made with the view of determining the weight to be ascribed to different experimental investigations. As to the seat of production of the intraocular fluid, nearly all authorities are agreed that it is produced by. the ciliary processes. From these processes a minute proportion travels backwards into the vitreous cavity, to be absorbed by the lymphatics of the optic disc, while by far the greater part makes its way between the lens and the ciliary processes, through the fibres of the suspensory ligament, into the posterior chamber, whence it passes round the margin of the iris into the anterior chamber. In addition to this mode of production, it has been suggested by Ehrlich that an appreciable amount of intraocular fluid may be secreted directly into the anterior chamber by the anterior surface of the iris. The experiments of Ehrlich(1) were made by the injection of a diffusible substance, fluorescine, aud we agree with Leber (2) in regarding them as proving the possibility of diffusion between the vessels in the iris and the anterior chamber, but not the secretion of a normal intra- ocular fluid by this channel. At any rate, any fluid formed in this way is negligible when compared with that which is produced in the neighbourhood of the ciliary processes. On the other hand, the place of absorption of the intraocular fluid is universally agreed to be the angle of the anterior chamber. Here the fluid is passed under pressure into the spaces of Fontana, whence it makes its way into the canal of Schlemm, between the endothelial cells lining this canal, and so is carried away into the venous system. This absorption is continuous, and its rapidity is largely determined by the height of the intraocular pressure. Since we have a constant absorption and a constant pouring out of fluid into the eyeball, it is evident that the intraocular pressure must be Factors which Determine Production of Intraocular Fluid. 295 a product of the two factors, formation and absorption, and that the main- tenance of the pressure at a constant height must be determined by an accurate balance between these two processes. The problem which lies before us is to determine the mechanism of formation of this fluid. The intraocular fluid is a clear, colourless solution containing a proportion of salts similar to that of the blood plasma, but having an osmotic pressure which is somewhat higher than the blood plasma, and containing the merest trace of proteids.* I. Methods of Research. The animals used were mostly cats. In a few cases dogs were employed, and in one experiment a rabbit. In the case of the cats the anesthetic used was always ether, with the addition in some cases of a small dose of morphia. In a few experiments, after the induction of full anesthesia, a small dose of curare was given. The administration of the anesthetic was continued during the experiment by an air-pump connected with a cannula in the trachea. For the dogs the A.C.E. mixture was employed. A record of the blood pressure was kept in all experiments. In some it was taken continuously, but in the greater number of experiments a short record was taken every few minutes in order to avoid trouble with clotting in the cannula. In the cats the blood pressure was taken in the lower part of the abdominal aorta, in the dog in the femoral artery. The apparatus we employed for measuring the intraocular pressure was very similar to that described in a former paper (3). A graduated tube with internal bore of about 0°5 mm.,and about 50 em. long, is provided with a lateral tube near each end. One end of the tube is connected by india- rubber tubing, by means of a T-piece, with a reservoir containing Ringer’s solution (or any other fluid the absorption of which is to be determined), and also with a manometer. The other end is connected by a second (glass) tube with a gilt steel hollow needle, which is introduced into the anterior chamber of the eye. The needle may be open at the end, or be closed at the end and provided with a lateral opening. To each of the side tubes a rubber capsule is attached. The capsule nearest the reservoir contains air, while that towards the eye is filled with fluid. By means of screw-clamps, fluid or air may be driven from either of the two capsules into the graduated tube. Before introducing the needle into the anterior chamber, the pressure in the apparatus is adjusted by raising the reservoir to about 25 cm. H20, which represents the average intraocular pressure. While the fluid is dropping from * Full details of various analyses of intraocular fluid are given by Leber (2), p. 207, et seq. 296 Mr. E. E. Henderson and Prof. E. H. Starling. [Nov. 23, the end of the needle, this latter is thrust through the lateral part of the cornea, so as to lie in the middle of the anterior chamber. A bubble of air is introduced into the graduated tube by compression of one capsule. and brought to the middle of the tube by relaxing the clamp on the capsule at the end towards the eye. The reservoir is then rapidly adjusted to such a height that the bubble remains stationary. In some of the later experiments a platino-iridium cannula, with a solid steel point made slightly conical, was found to be an improvement, as, in the event of any leaking occurring, 1t could be pushed in further. In introducing the cannula great care must be used, as, should the needle catch in or tear the iris, or wound the lens, the eye would be rendered useless for the purposes of the experiment. The needle, being comparatively large and blunt, requires considerable force for its introduction. We have found it safer to make a small perforation with the point of a cataract knife and, without letting the aqueous humour escape, to introduce the cannula in the hole thus made. Should the exact spot be lost sight of, a little fluorescine will stain it. A fine silk thread passed through the episcleral tissue, as in the operation for advancement of a rectus tendon, gives a better hold than fixation forceps, and is somewhat less in the way. The fluid employed in the apparatus was usually Ringer’s solution, in some cases normal saline. Whichever fluid was employed, it was filtered through a Berkefeld candle before the experiment, in order that no foreign body might be present which could lodge in and block the filtration channels. The intraocular fluid must play a twofold function in the eye. In the first place, by keeping up the intraocular pressure, it lends rigidity to the supporting structures of the eyeball, and furnishes therefore a fixed point for the intraocular muscles to contract against, besides maintaining the proper distances between the various refractive media. In the second place, it is the only source of nourishment to certain of the structures of the eye, namely, the middle and back part of the cornea, the lens and suspensory ligament, and the vitreous humour. The question that we have to decide is whether this fluid is formed by a process of secretion by the cells covering the ciliary processes, or whether it is a transudation similar to lymph. The question presents many analogies to that with regard to the secretion of urine. In each case we have a possible source of transudation in the capillary blood-vessel network and also an absorbing mechanism. We can only arrive at a conclusion by determining the physiological conditions under which we may alter either the production or the absorption of the intraocular fluid. 1905. | On the Production of Intraocular Fluid. 297 Il. The Effect of Changes in the Circulation on the Formation of Intraocular Fluid. If the production of intraocular fluid is dependent on a process of filtra- tion through the blood vessels and the epithelium covering the ciliary processes, its rate must vary directly with the difference of pressure on the two sides of the filtering membrane. It must vary, therefore, directly with changes in the capillary blood pressure, and inversely with the changes in the intraocular pressure. In our first series of experiments we sought to eliminate the second factor, namely, that of absorption, by opening the anterior chamber, so that the intraocular pressure could be regarded as zero. A cannula was introduced into the anterior chamber and the fluid allowed to flow off into weighed porcelain capsules. These were changed every 10 or 20 minutes, and the amount of fluid secreted in the time determined by weighing. The fluid drained off during the first minute after insertion of the cannula was regarded as normal intraocular fluid, but the gradual emptying of the eye-ball continues during the first five minutes, so that the figures obtained during this time cannot be regarded as expressing the rate of secretion. In every case the total solids of the intraocular fluid were also determined. The following experiment, p. 298, shows the results obtained while the blood pressure was approximately constant. It will be seen that there is a constant diminution in the amount of fluid obtained. In these experiments we were at first troubled by the formation in the anterior chamber of clots, which tended to plug the cannula. We found that this difficulty could be obviated by the injection of a dose of leech extract, not large enough to cause a permanent diminution of the blood pressure.* The next question to determine was whether it was possible to alter the rate of production or the composition of the intraocular fluid by altering the blood pressure in the vessels of the eye-ball. The experiments on this point were all carried out on dogs. A diminution of the intraocular blood pressure was easily effected by ligature or obstruction of the carotid artery on the same side. In order to produce a maximal rise of pressure in the blood-vessels of one eye, the vertebral and subclavian arteries on both sides were tied. A loose ligature was placed round the thoracic aorta, so as to permit of its being obstructed at any given time. A cannula connected to the mercurial manometer was placed in the carotid artery on the right side. The production of intraocular fluid was determined in the left eye. By obstruction of the * This procedure had been previously employed by Mr. E. Pfliiger in some experiments carried out in this laboratory in 1902. An account of these experiments will shortly appear.as a dissertation in the University of Bern. 298 Mr. E. E. Henderson and Prof. E. H. Starling. [Nov. 238, aorta a large rise of blood pressure was produced in this eye, since all the blood had to pass through the one carotid artery in order to get back to the heart. On the other hand, an almost complete anemia could be produced in the eye by obstruction of the one remaining carotid. We give below the results of one such experiment. Cat, anesthetised with Ether and the A.C.E. Mixture. A small dose of Curare was injected after anesthesia was complete. The extract of 2 grammes of dried leech heads was injected. | Weight of ; c lids aft Time. B.P. » mm.| Weight of are Percentage | Rate of flow g. secretion. Borisbanit of solids. | per minute. | weight. grammes. | grammes. grammes. 3.50 cannula inserted. Bai ososntesaossoosuanaeos ~ 130 0-689 0-009 1°3 BHAT essd00 594 egae09000800500 145 0:252 | 0-007 2-7 0:05 GBI) 4a: nanSBaceueeoose 120 0-756 | 0°032 4,°2 0-037 A BG Ns jrctuee weasanen denise 100 0-475 0-021 44 0-023 ANS Greatecetccmesecceceoscses 96 0 482 0-024 4°9 0-024 Dog. Weight, 73 kilos. Anzsthetised with the A.C.E. mixture and morphia. The extract of 2 grammes of dried leech heads was injected. Both subclavians and vertebrals were tied. Temporary ligature round aorta. Cannula in left eye. 3B.P. observed in right carotid. - Amount of Total Time. oe : secretion in | solids in eecntsee pra Remarks. grammes. | grammes. 3°29 —_ — — — — Cannula inserted. Aorta unobstructed 3°30 110 0°811 0-013 1°5 — 3°35 110 0 *432 0°014 3:2 “086 3°45 100 0 ‘550 0 027 4:9 0-055 BEDD 205 1°153 0-068 5°9 0-115 | Aorta obstructed. Fluid tinged red. 4°5 100 0 ‘627 0 ‘039 6:2 0-062 | Aorta unobstructed. 4°15 198 0°816 0-053 6°6 0-081 | Aorta obstructed. It will be seen that in every case a rise of intraocular pressure caused an increase in the amount of fluid secreted. It is impossible, however, to deduce directly from these experiments that the intraocular fluid is a transudation. The opening of the eye-ball and the consequent diminution of the intraocular pressure to nothing have a serious effect on all the intraocular structures. 1905. | On the Production of Intraocular Fluid. 299 Great dilatation of the vessels of the ciliary processes and iris is produced. The fluid, which, in the normal eye, is free from fibrinogen and contains the merest trace of proteid, rapidly acquires the power of coagulation, and its proteid content rises to 3, 4, or 5 per cent. The serious alteration of the vascular structures is shown in many cases by the appearance of red blood corpuscles in the fluid dropping from the cannula, and Greeff has shown that if the lowered pressure be brought about suddenly and maintained for some time, the epithelium covering the ciliary processes may be raised from the surrounding tissue so as to form small blisters, which are filled with coagul- able lymph. It has been suggested by Greeff (4) that the change in composi- tion of the intraocular fluid ensuing on opening the eye-ball is determined by the separation of the epithelium, but Bauer (5) has shown that the proteid contents may be raised in the absence of these epithelial changes, and that, on the other hand, the epithelial changes may be well marked on the subsequent day, when the wound in the cornea has closed, and the intra- ocular fluid has regained its normal composition. He also points out that the amount of change produced depends entirely on the rapidity with which the intraocular pressure is lowered. The change in composition is probably due, as Leber suggests, to the great distension of the capillaries and the consequent separation of their endothelial cells. It represents in fact an alteration in permeability of the filtering membrane. II. Amount of Intraocular Fluid Produced under Normal Circumstances. Im any investigation of the factors determining the production and absorp- fion of intraocular fluid, it is important to get some idea of the amount of this fluid secreted under normal circumstances, that is at normal intraocular pressure, Since the intraocular pressure is maintained constant so long as the blood pressure is steady, the amount of fluid produced at a given intra- ocular pressure must be equal to the amount of fluid absorbed at the same pressure. It is therefore a matter of indifference whether we measure the amount formed or the amount absorbed at any given pressure. Le Plat (6) sought to abolish the absorption of the intraocular fluid by filling the anterior chamber with oil or vaseline. A cannula was placed in the vitreous cavity, and the pressure in the cannula maintained at the normal intraocular pressure. It was found that the obstruction of the absorbing angle of the eye-ball carried out in this way caused a rise of intraocular pressure if the eye-ball were closed, or a flow outwards of intraocular fluid by the cannula if the pressure in this was maintained at the normal intraocular pressure. The amount of this outflow was measured, and was regarded by Le Plat as representing the normal rate of formation of intraocular fluid. He arrived at VOU) Evil: ¥ 300 Mr. E. E. Henderson and Prof. E. H. Starling. [Nov. 23 the conclusion that the amount of fluid normally secreted by the ciliary processes is in the rabbit about 4 c.mm. per minute. We found considerable difficulties in applying this method, chiefly determined by the tendency of the cannula in the vitreous to become blocked. We therefore adopted a method similar to that already employed by Niesnamoff, (7) under Leber’s direction. The arrangement of the experiment was as follows :— The hollow needle, connected by the capillary tube (containing an air bubble as index) to the reservoir and manometer, was introduced into the anterior chamber. The height of the reservoir was then adjusted until the bubble was stationary, showing that the intraocular pressure was exactly balanced by the pressure of the fluid in the tube leading to the reservoir. This intraocular pressure was of course maintained by a constant secretion of intraocular fluid, exactly equal to the amount escaping by filtration through the anterior angle of the eye. The animal was then killed by dividing the heart. This procedure at once stopped the production of intraocular fluid. The intraocular pressure, however, was maintained at its previous height by the connection of the eye with the reservoir of Ringer's fluid; the escape fluid by the anterior angle was therefore the same as before. The rate of this escape could be determined by noting the rapidity with which the air bubble moved along the capillary tube towards the eye, and this rate must be equal to the rate of production of fluid previously obtaining in the eye under normal conditions of circulation. The following table gives the rate of production of intraocular fluid, determined in this way, with varying intraocular pressures :— Intraocular Inflow, after cessation of Animal, pressure in mm. circulation, in cubic Hg. millimetres per minute. (Oy repyacdooucopacdeds 20 12 Cater Aicdamieaine: 15 1l (CE bash. cosemboannadose 26 12 (Ohieoonaoesseqnabdand 28 10 (Giitpasseacosesanagsa 14 5 (Ohh regannpenpancacsn 20 15 Average...... 20°35 10°8 It will be seen that there is a considerable difference in the case of filtra- tion in various eyes, and therefore a corresponding difference in rate of production of intraocular fluid, 1905. | On the Production of Intraocular Fluid. 301 IV. The Factors Determining Absorption of Intraocular Fluid. In the last set of experiments we determined the rate of absorption of intraocular fluid at the normal intraocular pressure, and regarded this as representing the rate of production of this fluid under normal circumstances. In the same experiment it was possible to alter the intraocular pressure by raising or lowering the reservoir, and so to determine the effect of the height of the intraocular pressure on the rate of absorption. The results of two such experiments are given below, and show conclusively that the rate of absorption is determined, in the absence of disturbing factors which we shall have to consider later on, solely by the height of intraocular pressure. (1) Cat, anesthetised with Ether. While the anesthesia was maintained, a small dose of morphia and curare was injected. Atropine was instilled locally into the conjunctival sac. Rate of inflow in cubic B.P. in mm. Hg. | 1.0.P. in mm. Hg. millimetres per minute. 115 22 10) 115 30 4 115 46 a 130 62 8 0 22 12 0) | 36 16 0 46 19 0 5 (2) Cat, aneesthetised with Ether. Atropine and cocaine instilled locally into the conjunctival sae. B.P. in mm. He. | 1.0.P.in mm. He. Hes @ aalllos tn OL millimetres per minute. 124 32 0) 124: 44. 5) 12 52 am 110 20 0) 116 Ad 10 116 52 20 Heart divided. 10) 52 22 0 44. 15 0 20 12 302 Mr. E. E. Henderson and Prof. E. H. Starling. [Nov. 23, In a previous paper we have shown that the intraocular pressure varies directly as the blood pressure in the vessels of the eyeball. We must therefore conclude that the rate of absorption of intraocular fluid is also determined by the height of the blood pressure, and since the absorption must keep pace exactly with the formation of this fluid, it follows that the formation of the intraocular fluid must also be determined by the height of the intraocular blood pressure. So far then the conditions which we laid down as necessary to be fulfilled in order to justify the filtration theory of the production of intraocular fluid have been fulfilled, and we might conclude with Leber that the formation of this fluid is exactly analogous to that of lymph, and is determined by the difference of pressure between the blood in the vessels and the fluid outside the vessels. There are, however, certain difficulties in this assumption which have so far not been considered by previous workers, but which must be met satisfactorily before we can come to any definite conclusion on the subject. Tt has hitherto been assumed by Leber, Niesnamoff, and others, that a fluid having the composition of intraocular fluid might be formed by a process of filtration through the blood vessels of the ciliary processes under any difference of pressure. In this assumption they have neglected the question of the different proteid content of blood plasma and intraocular fluid. It was shown by one of us (EK. H. 8.) that, in order to separate a proteid-free transudate from a fluid such as blood serum, a certain amount of work had to be done, and that for this separation a minimum difference of pressure on the two sides of the filtering membrane of at least 28 mm. Hg was necessary. The intraocular fluid has such a small content in proteid that it may be regarded as analogous in all respects to the fluid which is supposed to be separated vy the glomeruli of the kidney. In order therefore that any fluid shall be poured out in the eyeball, a minimum difference of 30 mm. Hg must be present between intraocular pressure and capillary blood pressure. If this pressure difference is not present, work must be done by the cells forming the filtering membrane, and the formation of intraocular fluid must be regarded in the light of a secretion rather than in that of a transudation. A definite decision on this point could be reached if we had any means of determining the blood pressure in the capillaries of the eyeball. A method for this purpose has been devised by Niesnamoff, (7) and this observer states that the normal intraocular capillary pressure is about 50 mm. of mercury. His arguments, however, involve several fallacies. In his experiments he con- nected a cannula, attached to a reservoir of salt solution, with the eyeball of a living animal. He found that the fluid neither ran in nor out at 25 mm. Hg, which was therefore the intraocular pressure. He then 1905. | On the Production of Intraocular Fiud. 303 determined the rate of inflow when the pressure in his cannula was raised to 50 mm., 75 mm., and 100 mm. He. He then killed the animal, and again determined the rate at which the fluid would flow in under these various pressures. He found that above 50 mm. Hg the rate of inflow was the same in the dead as in the living animal. He therefore concluded that 50 mm. Hg represented the intracapillary pressure. In coming to this conclusion he was guided by the assumption that, when the intraocular pressure was raised so as to be equal to the intracapillary pressure, the transudation of intraocular fluid would cease, and above this pressure the rate of inflow for his reservoir would be, therefore, the same in the living and dead eye. It is impossible, however, by this method to determine intracapillary pressure. The globe of the eyeball is practically mgid. As the intraocular pressure is raised, the intraocular fluid will press upon the veins of the ciliary processes, and the blood pressure will therefore rise in the capillaries and in the veins until it is greater than the intraocular pressure. With successive rises in the intraocular pressure the pressure in capillaries and veins must get larger and larger in order that any circulation of blood may be maintained, and the circulation through the capillaries will cease only when the intraocular pressure is very nearly as high as the arterial pressure. If the circu- lation in Niesnamoff’s experiments ceased at 50 mm. Hg, it is evident that the normal intracapillary pressure, when the intraocular pressure is 25 mm. Hg, must be considerably below 50 mm. Hg. How then are we to explain the very definite figures obtained by Niesnamoff? This observer apparently performed very few experiments. In his paper he gives the results of only one such experiment as that here described. On repeating his experiments we found it impossible to obtain anything like such definite figures—and this for various reasons. In the first place, a considerable rise of intraocular pressure, such as to 50 or 70 mm. Hg, exercises an abnormal stretching effect upon the filtering apparatus of the eyeball, so that the channels at the anterior angle of the eye are gradually opened up, and in many experiments we observed a consequent gradual increase in the rate ot inflow of the fluid. In most experiments, for example, the rate of inflow was greater with descending pressures than with ascending pressures. This is well shown in experiment No. 2, on p. 301. The following experiment shows the dilatation consequent on a preliminary raising of the intraocular pressure :— 304 Mr. E. E. Henderson and Prof. E. H. Starling. [Nov. 238, Cat, anzesthetised with Ether. serine applied locally to conjunctival sac. Pupil moderately contracted. BP. in mm, Hg, | 1L.0.P. in mm. ag. eee eee eta | 110 | 16 ) 110 | 32 5 110 | 48 8 108 | 64 9 112 16 0) 112 32 § 112 48 13 | 112 64 18 Another disturbing factor is the size of the pupil. We shall have to consider this factor more in detail later on, but unless atropin be given at the beginning of the experiment, the observations on the living eye are made with a somewhat contracted pupil, whereas those on the dead eye are made on a widely dilated pupil. Other factors being equal, the filtration in the eye with dilated pupil is always slower than in the eye with contracted pupil. In certain of our experiments we observed an equality of inflow between the dead and living eye at some pressure above 40 mm. of mercury, but on further raising the pressure this equality disappeared, showing that we were dealing with yielding tissues and altering membranes. This fact rendered it impossible to obtain by such methods any definite information of the intracapillary pressure in the eye-ball, or of the level of intraocular pressure at which transudation or formation of intra- ocular fluid would definitely cease. One other factor which would aid in disturbing the results obtained is the effect of a high intraocular pressure on the general circulation through the eye-ball. If we succeed in raising the pressure to such a height that the circulation is entirely abolished, changes must rapidly take place in the apparatus both for formation and absorption of intraocular fluid, and subsequent results cannot be compared with those obtained before such a cessation of circulation. The raising of the intraocular pressure in itself may act as a stimulus and cause reflexly alterations in blood flow, in the general blood pressure, or in the state of contraction of the pupil. The co-operation of these various factors suttices to explain the varying results obtained in the very many experiments we performed upon this subject, including those of which we have already given details. We are of opinion, therefore, that the results obtained by Niesnamoff must be regarded as accidental, and that a greater number of experiments would have convinced this observer of the fallacies of his method. 1905. | On the Production of Intraocular Flind, 305 Although it is impossible at present to determine the intracapillary pressure in the ciliary processes, we may at avy rate inquire whether there is, in all experiments on the subject, the possibility of a difference of pressure of 30 mm. He between intracapillary blood pressure and intraocular pressure. In the case of a similar question in the kidney, we have been accustomed to compare the aortic blood pressure with the ureter pressure, and have regarded a difference of 40 mm. between these two pressures as satisfying the necessary conditions for filtration through the glomeruli. A similar comparison of arterial blood pressure and intraocular pressure leads to the same result. Below we give the intraocular pressure and arterial pressure as determined in a series of 20 experiments. It will be seen that in every case there is a difference between the two pressures of at least 48 mm. Hg, the average difference of pressure in all the experiments being 84:8 mm. He. =e : Animal. BE ‘He yates a eaees Lae eps sete sb g. Ife. Caltinscaenehiecsnccrns 1380 16 114 Gatien cdastises 140 25 115 (ODirocaseacesusostonde 188 20 118 Cate atm Asce snes 94 24 70 IRON Sonoposodone 74 16 58 ID ays: topdagaceseeee 112 14 98 Gath eae a des cbs 104 15 89 Watiecescaciaascaees 106 19 87 Cat 106 18 88 | Cath tieschicssaecasien 120 20 100 (OD Hbchase fae Sap seReod 150 22 128 IDL Coanonceouaance 84. 12 72 Dog 58 10 48 Doge ered 70 16 54 (Ob) ibecoecodcesbnaded 115 23 92 Caterers sionen 12 32 92 Gata deve eceaet ca 110 16 94, Catissanccneessnce nse 138 22 116 Cate aaiiae comic 94 27 67 Catisanasaacgeoueaeeceh 110 24 96 So far then our observations tend to support in every particular the view laid down by Leber, namely, that intraocular fluid is produced in the ciliary processes by a process of filtration, and that the sole factor determining the amount of transuded fluid is the difference of pressure between the blood in the capillaries and the fluid in the eye-ball. V. Influence of the Proteid Content of the Intraocular Kluid on the Intraocular Pressure. The fact that the intraocular fluid has to be filtered through the intercellular channels of the endothelium bounding the spaces of Fontana and lining the 306 Mr. E. BE. Henderson and Prof. E. H. Starling. [Nov. 23, canal of Schlemm, in order to escape from the eyeball, suggests that the resistance will be greater if the viscosity of the filtering fluid be increased in consequence of raised proteid content. Indeed, one form of raised intra- ocular pressure, the glaucoma accompanying inflammation of the ciliary region, has been ascribed to the greater proteid content of the intraocular fluid secreted by the inflamed vessels, and the consequent greater resistance to the filtration of this fluid through the anterior angle of the eye. So far as we are aware, there are no direct determinations of the relative rates of filtration of normal salt solutions with and without proteid. We have, therefore, in a series of animals, determined the intraocular pressure under the two conditions :— (a) With normal intraocular fluid. (b) After replacing this fluid by blood serum. We have also compared the relative rates of filtration of normal salt. solution and of serum in the living and dead eye. In our experiments one eye of the animal was connected with a reservoir and manometer containing Ringer’s saline fluid, while the other was connected with a similar apparatus filled with filtered blood serum. In order to determine the intraocular pressure in an eye, in which the normal aqueous humour had been replaced by serum, after introduction of the hollow needle, the aqueous was allowed to escape through the side opening in the cannula. Serum was then allowed to flow in for a time, and then the contents of the anterior chamber again allowed to escape. The side tube was then closed, an air bubble introduced into the capillary tube, and the pressure determined at which the bubble moved neither backwards nor forwards. In nearly every experiment the intraocular pressure, during the first 5 or 10 minutes after the insertion of the cannuia, was higher in the eye filled with serum than in the eye filled with normal fluid. The difference, however, rapidly diminished, so that 15 to 20 minutes after the beginning of the observation the pressures were practically identical in the two eyes, and remained so throughout the rest of the experiment. It must be remembered that with the zero method used by us there is no movement of fluid into the eye. Hence the fluid necessary to replace the loss by filtration and to maintain the intraocular pressure is being constantly secreted by the ciliary processes, and is probably of the normal composition, 7.c., practically free from proteid. We should therefore expect a gradual decline of the intra- ocular pressure in the eye with serum, although hardly so rapid an equalisation of the pressures on the two sides as we actually observed in our experiments, 1905. | On the Production of Intraocular Fluid. 307 After the determination of the intraocular pressure, the animal was killed by opening its heart, and the inflow of serum and saline fluid respectively observed, first under the normal intraocular pressure, and then under raised pressures. The results of two such experiments are given below. It will be seen that there is a marked difference in the rate of filtration of the two fluids, that of serum being, as one might predict, very much slower than that of saline.* Experiment 1—Dog, A.C.E. Morphia. Curare. _ Vagi cut. | Intraocular pressure. Time. | Blood pressure. ——— = | Salt eye. Serum eye. 4.15 70 mm. Hg. | 26 °2 29 -4 em. water. 4.20 70 i | 24-2 27 5 445 100 5 29:2 29 5 Animal lulled | by opening heart. Inflow per minute in cubie milli- metres (after 10 minutes). Pressure. | | ] | Salt. Serum. | 29 cm. | 11°5 6 oa 1a 59 } 6 | — | 1:5 | 6 Experiment 2.—Cat. Ether, morphia, curare. | | Intraocular pressure. Time. Blood pressure. or ta j | | Salt. Serum. | ' ar . 3.0P.M. | 120 148 15-1 S10 116 10°8 125 ly si20 | 110 9°2 me | Animal Killed. | * Although serum filters more slowly than normal intraocular fluid or saline, the difference is not sufficiently great to cause any marked variation in the intraocular pressures on the two sides. One cannot, therefore, in view of these observations, ascribe any large part in the production of any form of glaucoma to possible differences in the composition of the aqueous humour which might be determined by inflammatory conditions of the blood vessels. VOL. LXXVII.—B. N ! 308 Mr. E. E..Henderson and Prof. E. H. Starling. [Nov. 23, Inflow three minutes later at same intraocular pressures— Salt. Serum. 6 3 5 3°5 5 6 i) 4 A 4 Fifteen minutes later— 3:5 15 30 15 Bi) 1G ete ete This difference in the rate of filtration of the two fluids becomes greater the higher the intraocular pressure is raised. VI. The Effect of the Size of the Pupil on the Absorption of Intraocular Fluid. In the experiments we made to decide this point, one eye of the animal under observation was treated with eserine and the other with atropine. The instillation of these drugs should be begun before the induction of anesthesia, as the action of eserine is very uncertain if only instilled after anzesthesia. We have found, as a result of these experiments, that the intraocular pressure in the two eyes remains the same during the time of observation, but that, if the pressure in the apparatus be raised, the rate of filtration in the eye under eserine is much greater than in that under atropine. It is difficult to give a precise explanation as to the cause of this difference. Stretching of the filtration spaces at the angle of the anterior chamber may possibly account for it all. If this, however, is the case, we should expect to find the intraocular pressure at a lower level in the eye with the contracted pupil, for the intraocular: pressure must of course be the product of the rate of secretion and the rate of absorption of the intraocular fluid. The same objection applies to the explanation of this phenomenon by Grénholm (9), who states that in his opinion it is due to diminished intraocular secretion as a result of the contraction of the intraocular vessels. It may also be possible that at these raised pressures other channels of filtration are opened up— such for instance as the surface of the iris. An important, perhaps the most important, factor, however, must be the crushing of the dilated flaccid iris 1905. | On the Production of Intraocular Fliad. 309 into the filtration angle of the eye, thus causing a mechanical obstruction, which will be more marked the greater the intraocular pressure. Hence the smaller amount of filtration in the atropinised or dead eye with dilated pupil, as compared with that in the eye which has been put under the influence of eserine. The figures of a typical experiment are given. Cat, anesthetised with Ether. Blood pressure average 138 mm. Hg, with only trifling variations throughout the experiment. Rate of filtration in Rate of filtration in | Rate of filtration in Intraocular pressure | eserine eye in cubic | atropine eye in cubic | atropine eye post- in mm. Hg. millimetres per millimetres per | mortem, in cubic minute. minute. | millimetres per minute. 20 6) 0 | 15 35 11 8 20 50 | 16 1 | 25 65 23 14. 31 | Summary of Conclusions. 1. The intraocular pressure represents the pressure at which the rate of formation of intraocular fluid is exactly balanced by its rate of escape through the filtration angle of the eye. 2. The production of intraocular fluid is strictly proportional to the difference of pressure between the blood in the capillaries of the eyeball and the intraocular fluid. 3. No satisfactory method of measuring the intracapillary pressure in the eyeball has been yet devised. The fallacies of Niesnamoff’s method are pointed out. Judging, however, from a comparison of the arterial pressures and the intraocular pressures in a large number of animals under different conditions, there is probably always a difference between the intracapillary pressure and intraocular pressure, which is sufficient to account for the production of the intraocular fluid, without assuming any active intervention on the part of the cells of the capillary walls or of the ciliary processes. 4. An increased proteid content of intraocular fluid slows its rate of absorption in consequence of the mechanical hindrance of the proteid to filtration. 5, Filtration, 2.e., the absorption of intraocular fluids, at high intraocular pressures is favoured by constriction of the pupil and hindered by dilatation of the pupil. The difference, however, is barely perceptible with normal or low intraocular pressures. 310 Production of Intraocular Fluid. The expenses of this research were defrayed by a grant from the Scientific Grants Committee of the British Medical Association. BIBLIOGRAPHY. 1. Ehrlich, ‘ Deutsche med. Wochenschr., No. 2, ff., 1882. 2. Leber, ‘ Graefe-Saemisch, Handbuch der Gesam. Augenheilkunde, vol. 2, pt. 2, 1903. 3. Henderson and Starling, ‘Journ. of Physiol.,’ vol. 31, pt. 3, 1904. 4. Greef, ‘ Arch. f. Augenheilk., vol. 28, pp. 176—192, 1894. 5. Bauer, H., ‘v. Graefe’s Arch. f. Ophth.,’ vol. 42, p. 3, 1896. 6. Lepat, ‘Ann. d’Ocul.,’ vol. 101, 1889. 7. Niesnamoff, ‘v. Graefe’s Arch. f. Ophth.,’ vol. 42, p. 4, 1896. 8 Starling, ‘Journ. of Physiol.,’ vol. 19, 1896, p. 312. 9. Gronholm, ‘y. Graefe’s Arch. f. Ophth.,’ vol. 49, 1900. dll On the Filtration of Crystalloids and Colloids through Gelatine : with special reference to the behaviour of Hamolysins. By J. A. Craw, British Medical Association Research Scholar. (Communicated by Leonard Hill, M.B., F.R.S. Received December 1, 1905,— Read February 1, 1906.) The current controversy between Ehrlich (1898, 1903) and Arrhenius and Madsen (1902, 1904) on the physical chemistry of the neutralisation of toxins by their specific antitoxins led the author (IV, 1905) to an investigation of the relations existing between the toxin for red blood corpuscles secreted by B. megatherium and its specific antiserum. One of the methods adopted consisted in the filtration of mixtures of megatheriuin hemolysin and antilysin through gelatine. The lysin was found to pass into and even through the gelatine, whereas the antilysin was retained, and by means of the delicate blood test for free lysin it was possible to demonstrate that the two substances, on mixing in any proportions, do not completely neutralise each other. These observations indicated that Ehrlich’s views on the toxin- antitoxin reaction required considerable modification, but a closer investigation showed that the hypcthesis advanced by Arrhenius and Madsen agreed even less with the experimental facts. On the other hand, the results were in entire harmony with the views advanced by Landsteiner (1903) and Bordet (1903), which have been supported by Nernst (1904) and Craw (1905) that the toxin is adsorbed by the antitoxin much as a dye is by a tissue. As this conclusion may considerably modify ‘current ideas on the nature of the reaction and the constitution of toxins in general, it seemed advisable to inquire further into the physical chemistry of filtration through gelatine. The present communication contains data of the filtration of various erystalloidal and colloidal solutions, including megatherium lysin, through various percentages of gelatine, under constant and variable pressures. The work was partly carried out as Research Student at the Lister Institute of Preventive Medicine and was completed at the London Hospital Medical College by the aid of a scholarship from the British Medical Association. Previous Work on the Gelatine Filter—The gelatine filter was introduced _ by C. J. Martin (1896), and consists of a Pasteur-Chamberland candle, the pores of which are filled with solid gelatine. The filter is fitted into a gunmetal jacket or filter case, which serves to hold the liquid to be filtered, and the upper end of the closed filter case is connected with a supply of air at a pressure of 30 to 100 atmospheres, which is used to force the liquid VOL. LXXVII.—B. : 2A 312 Mr. J. A. Craw. On the Filtration of [ Dec. 1, through the gelatine. From his observations with “wet” gelatine filters, a.¢., filters containing gelatine from which the normal content of water had not been removed by drying in air or otherwise, Martin* concluded that gelatine was impermeable to colloidal substances such as globulin, albumin, glycogen, and soluble starch, but partially permeable to albumoses and dextrin, and completely permeable to solutions of crystalloids, eg., urea and dextrose. It seemed, then, as if the gelatine filter was an instrument destined to play an important part in the investigation of physiological fluids. Martin and Cherry (1898) applied the filter to the investigation of the course of the reaction occurring between diphtheria toxin and antitoxin, and likewise to the reaction of snake venom with antivenene, the toxin and venom being filter-passers, whereas the anti-bodiest+ were retained. From these experiments it seemed as if the toxin was completely neutralised by the antitoxin, but further investigation of the mechanism of gelatine filtration shows that no such absolute conclusion can be drawn. E. Waymouth Reid (1901) showed conclusively that crystalloids do not pass “wet” gelatine filters in unaltered concentration, and that although filters which had been dried to constant weight in dry air allowed certain crystalloidal solutions to pass unchanged, yet dextrose and sodium oleatet were partially retained. Further, he found that the filtrate from serum} had not the same composition as the proteid-free serum, and that the residual fluid left in the filter case had a much higher concentration of organic substances of non- proteid character than either the original serum or the filtrate. The Martin filter is not, therefore, a simple means of separating crystalloids from colloids. One must not, however, under-estimate its value as an instrument for the analysis of physiological fluids, for although the filter shows considerable differences in permeability to various ecrystalloidal substances, I find (IV, 1905) that these and inferior colloids are, on the whole, retained to a small extent compared with typical colloids. The partial retention of filter-passers has an important bearing on the conclusions to be drawn from gelatine filtrations, for, if the concentration of the filter-passer to be tested for be small in the original fluid introduced into the filter, the gelatine may retain practically the whole amount and the filtrate contain only a quantity below the experimental error of observation. This was found to be the case for neutral mixtures of megatherium lysin and antilysin, and mixtures * Loe. cit. + Of. also Brodie (1897) (1900). { Krafft (1902) considers soap solutions, such as‘sodium oleate, to be colloidal. § Cf. also Starling (1899). 1905. | Crystalloids and Colloids through Gelatine. 313 containing excess of antilysin, by the author;* the filtrates showed no hemolytic power, whereas the gelatine had stored up a considerable amount of free megatherium lysin. In this light Martin and Cherry’s observations on diphtheria toxin and snake venom, mentioned above, are in entire agree- ment with my results for megatherium lysin, the free toxin and venom of the neutral mixtures being probably stored up in the gelatine of their filters. On “ Wet” and “Dry” Gelatine Filters—Martin and others used the filters in the “wet” state, zc. shortly after solidification of the gelatine in the pores of the candle, and, therefore, containing a considerable quantity of water, part of which on filtration will pass into the filtrate and so dilute the substance filtered. To get rid of this difficulty, E. W. Reid removed part of the water by drying Martin candles to constant weight in a current of dry air. It seemed to me, however, that by drying the gelatine another difficulty might be introduced, viz., a change in the size of the pores, which would render observations with “dry” filters of doubtful value. Method.—The rate of filtration of water through freshly prepared wet filters was observed during one hour and compared with the rates during a similar period, of filters which had been partially dried by standing in dry air for 10 hours, 24 hours, and 3 days. Rates of Filtration—The freshly prepared wet filters gave fairly uniform rates, and allowed from 0°5 to 2 c.c. of water to pass per minute at 100 atmospheres pressure, the concentration of the gelatine in the pores being 9 per cent. and the temperature 10° to 15° C. Under the same conditions the partially dried filters allowed from 5 to 20 ec. to pass per minute at the beginning of the filtration, but the rate rapidly decreased. On reducing the applied pressure to about one atmosphere, and allowing the water to flow through a partially dried filter at the rate of about 1 to 2 c.c. per minute, the filter gradually tightened, so much that after the passage of 30 to 50 cc. a pressure of 100 atmospheres was necessary to maintain the same rate. The longer the filters were dried the more marked was the porosity and the greater the amount of fluid which had to be pressed through before the filter tightened. Conclusions.—On partially drying gelatine filters the gelatine shrinks, and air passages are produced of greater diameter than the water passages of a wet filter. The wide dry filter passages at first offer a free passage to the fluid filtered, and no material change is to be expected in the percentage composition of the latter. The filter gradually changes in character until ultimately, a wet filter is obtained with pores of similar dimensions to those of a freshly prepared wet filter. The various fractions of filtrate from a * Loe. ett. ey IN 314 Mr. J. A. Craw. On the Filtration of [ Dec. 1, “dried” filter are not subject to the same conditions of filtration and are not, therefore, comparable. Wet filters, on the other hand, show a much greater constancy in rate of filtration, and are, therefore, to be preferred. On the Preparation of the Fulters—To overcome the difficulty of dilution of the filtrates by the water of wet filters, the gelatine to be used in filling the candles was in several cases dissolved in the fluid to be filtered. Thus the filtration of 0°8 per cent. sodium chloride and of megatherium lysin took place through 9 per cent. gelatine which had been dissolved in 08 per cent. sodium chloride, and 1:54 per cent. potassium iodide was filtered through 11 per cent. gelatine in 1°54 per cent. iodide. A Pasteur- Chamberland candle of size B, which had in the majority of cases been heated for 10 hours at least in a muffle furnace to remove organic matter, was fixed into a brass socket with “Faraday” cement and fitted con- centrically into the internally tinned gunmetal jacket or “filter case.” After thoroughly washing through with about 400 cc. of hot water and 250 cc. of the solution of French gelatine to be used in forming the filter, at a temperature of 37° C., the gelatine at 30° C. was slowly passed through the cooling filter at an air pressure just sufficient to cause about 1°5 c.c. or 15 drops to filter per minute. The filtration was continued until no trace of air bubbles was visible in the drops, after which a further 50 c.c. were allowed to pass. During the final filtration the upper surface of the gelatine solution in the filter case was not allowed to descend below a level 3 cm. above the crown of the candle. The filter was then removed, drained, placed in the neck of a flask containing a little water, and kept in the ice chest at 6° to 8° C. until required. If the filter so prepared had an obviously thick skin of gelatine upon its surface, this was removed by immersing in the gelatine solution at 30° C. and draining as before. In the comparative experiments with different concentrations of gelatine, the filters were treated in the manner last described to get membranes of approxi- mately equal thickness. Only those filters were used which allowed less than 2 ¢.c. or about 20 drops per minute to pass at 100 atmospheres pressure and 10° to 15° C., and the majority of filters allowed only about 0°5 c.c. per minute to pass under these conditions. After use the candles could be washed out with water at 50° C., dried, heated in a muffle furnace, and on refilling with gelatine again gave reliable filters. Filtration of a Typical Crystalloid, Sodium Chloride, through 9 per cent. Gelatine. Sodium chloride was chosen on account of its characteristic crystalloidal properties and the important réle it plays in physiological fluids. Further, 1905. | Crystalloids and Colloids through Gelatine. 315 in a previous communication,* it was found that the filtrates from megatherium lysin were strongly hemolytic, and it was necessary to determine if this hemolysis was due in any degree to change in the concentration of sodium chloride. Method.—The filters used contained 9 per cent. gelatine in 0°8 per cent. NaCl, and about 50 ec. of 0°8 per cent. NaCl were pressed through at 100 atmospheres and 10° to 12° C., the filtrates being collected in fractions of about 4 ec. The NaCl content of 1 cc. of the fractions of filtrate, the residual liquid left in the filter case and the original fluid, was determined by titration with 1/100 normal silver nitrate, using potassium chromate as an indicator. The hemolytic powers of the same fluids were determined in these and all other experiments in this paper, unless otherwise stated, by mixing 1 cc. of the respective fractions with 2 cc. of 2°5 per cent. euinea-pig’s red corpuscles which had been washed and suspended in 0°8 per cent. NaCl. The mixtures were heated to 37° C. for three hours, well shaken every 30 minutes, and allowed to sediment generally about 12 hours in the ice chest at 6° to 8° C. The intensity of tint of the supernatant fluid was then determined by comparison with the scale of a von Fleischl hemoglobinometer which had been standardised by a blood solution of known content. Complete heemolysis is indicated by the index (100). Hiaminatiion for Sodiwm Chloride and Hemolytic Power.—Experiment No. 2, Table I, represents one out of four similar experiments. Original Fluid: 0°81 per cent. NaCl given in the table as Orig. (100). Filtrates: The relative concentrations of the Ist, 4th, and 14th fractions were (47'1), (89°5), and (99-2) respectively. Residual Fluid: When tested immediately after decompression had the value (100) exactly, but, on allowing to stand 12 hours in the filter case, showed the relative concentration (105). Hemolytic Powers: No hemolysis was obtained under the standard conditions with either filtrates, residual fluid, or the original saline. Conclusions—The typical crystalloid, sodium chloride, is markedly retained on filtration through gelatine which originally contained the same concentration of saline as the fluid filtered. The salt taken up by the gelatine is expressed or diffuses into the residual fluid after decompression. The diminution in salt concentration in the filtrate is insufficient to cause hemolysis under the standard conditions. * Loe. cit. 316 Mr. J. A. Craw. On the Filtration of [Dee. 1, Continued Filtration of Sodiwm Chloride through 9 per cent. Grelatine. As the last fraction of filtrate in No. 2 did not quite reach (100), it seemed desirable to ascertain whether on continued filtration this value could be obtained. Method.—The residual fluid of No. 2, which had stood in the filter case Table I. Sodium chloride, per cent. | Megatherium lysin, H.Is. No. ... 1. 2. 2A | 2B. | 3 4 5 6 le | 8 1Es © 0°88. 0°81. | HBu. | 15 11. 9. 7°5 8. 23 °9 47 “1 101 °5 | 97°4.| 16-7 | 0:21 0-7 | 0°5| 0-0) 0-5 Oro 5 31 °4 55 ‘1 103°3 | 97°8 | 41°7.| 5:8) 05) 00; O00} 50| 3°5 8 39 0 76°1 99°6 | 96:0 | 47°8 | 60-4) O00 |] 2:3] 0:0)17-7 10-7 £ | 49:0 | 89°5 | 101-1 | 93:1] 56:9] 74:0] 0:0 | 5:1] 3-7 | 25-0 | 178 =| 56 °0 95 °3 99°6 | 94:2 | 65°6 | 85:4] O'°0 | 5:1] 27°8 | 34°6 | 21°4 os 64 +1 96 °7 97°6 | 93°5 | 70°6 | 88:5 | 0:0] 7:1 | 46°3 | 45:1 | 25°0 a 71:1 97 6 100°7 | 95:7 | 76:8 | 88:5 | 0:0} 8:6} 33°9 | 48°8 | 28°5 & 761 97 °6 97°6 | 94:2 | 80:4] 90'°6| 00) 8°6| 47°8 | 48°8 | 38°5 3 79 °2 97 6 = 94:2 | 88:1 | 90°6 | 0:0 | 8°6 | 47:0 | 48°8 | 32:1 & 81:7 99 -2 — 96:0 | 85:0 | 90°6 | 0:0 | 12°38 | 47-0 | 48-4 | 32:1 a 83 °3 99 °6 —_— 94°9 | 86°3 | 90°6 | 0O°0 | 12°8 | 45°6 | 48°4 | 35-7 i 112°6 | 100-0 —_ 96°8 | — _ 2°3 | 8-0 | 47:0 | 49°3 | 35°7 im 107 ‘8 99 -6 — 96°4 | — — 3°7 | 10:0 | 46°3 | 48°8 | 50-0 8 107 ‘8 99 °2 _ 96°4 | — — |100; — — | 48-4 | 50°0 a 98-4 = — | 96-4 —}|—}]—|— | — | 50:0 96 -2 — — — — — —_— _— — — | 50°0 93 -7 1384°9 | 100:0 | 100°0 |100°0 110 0 |125-0 | 85:0 | 62:9 | 55°6 | 52-7 | 67°8 Res. : C.c. c.c. C.c. cle.) eke: c.c c.c, | ce C.c c.c 20 44, 65 40 45 _ AT 45 34 31 100‘0 | 100°0 | 100°0 /100-0 |100°0 |100°0 | 50°0 | 50°0 | 50°0 | 50°0 | 50°0 Orig: C.c. C.c. C.c. C.c. | ¢.c. Ca | G@ | Ga | aa | Ge. 100 112 112 120 105 = 110 105 100 110 105 overnight, was poured out, and the filter case refilled with 112 cc. of saline, of which 40 c.c. in 5 cc. fractions were pressed through. Hxamination for Sodium Chloride—Experiment No. 24, Table I. Original Fluid: 0°81 per cent. NaCl = Orig. (100). Filtrates: The 1st and 2nd fractions contained (101°5) and (103°3) respec- tively, and those succeeding showed values varying about (100). 1905. | Crystalloids and Colloids through Grelatine. 317 Residual Fluid: Titrated immediately after decompression (100), after 12 hours in the filter case (110). Conclusion—On renewing filtration the filtrate may contain a higher percentage of salt than the original fluid, due probably to the temporary decompression, for on continued recompression the concentration tends to become less than (100). Continued Filtration of Sodium Chioride with Megatherium Lysin and Antilysin through 9 per cent. Gelatine. In a former paper,* mixtures of lysin and antilysin were pressed through gelatine which had been tested for tightness by the filtration of saline. ‘he redistribution of salt under these circumstances was now determined. Method.—A filter through which 10 c.c. of 0°81 per cent. NaCl had been pressed, the original gelatine being 9 per cent. in 0°81 per cent. NaCl, was decompressed, the residual salt solution removed, and the filter case refilled with 120 c@c. of a nearly neutral mixture of equal volumes of a fluid megatherium lysin and 5 per cent. antilysin in saline, of which 75 ce. in 5 c.c. fractions were pressed through. Hzamination for Sodium Chloride—Experiment No. 28, Table I, is one of two practically identical filtrations. Original Fluid: Contained 0°81 per cent. NaCl = (100). Hemolytic Index (16°8). Filtrates: As in No. 2A the 1st and 2nd fractions were the most con- centrated in NaCl, but they did not reach (100), and the final fraction was (96:4). The average hemolytic index of the whole filtrate was (0°5). Residual Fluid: The results were entirely similar to those of No. 2a as regards NaCl. Hemolytic Index (38'6). Conclusion —tThe redistribution of salt is insufficient to give a trace of hemolysis under standard conditions. Temporary decompression even for a few minutes allows a higher per- centage of salt to pass into the filtrate than would be obtained by constant pressure. Filtration of Sodiwm Chloride with Butyric Acid through 9 per cent. ; Gelatine. Butyric acid in saline acts as a very strdng heemolysin, and, as it possesses erystalloidal properties, it appeared to be of interest to compare its filtration phenomena with those of megatherium lysin in saline. Further, butyric * Loe. cit. 318 Mr. J. A. Craw. On the Filtration of [ Dee. 1, acid greatly diminishes the surface tension of saline against air, and the possibility of a change in the surface forces between saline and gelatine formed a second point of interest, which is discussed towards the end of this paper. Method—A_ solution of 0°81 per cent. NaCl containing 48 per cent. butyric acid was pressed through 9 per cent. gelatine in 0°81 per cent. NaCl at 100 atmospheres. The content of the fractions of filtrate, etc., in NaCl was determined as before, and the butyric acid was estimated by titration with standard sodium hydrate, using phenol-phthalein as indicator. The hemolytic powers of the various fluids were also determined by the time taken for 1 c.c. of the fractions to hemolyse completely 2 cc. of a 2°5 per cent. suspension of guinea-pig’s corpuscles in 0°8 per cent. NaCl at 16° C. Hxamination for Sodiwm Chloride—HExperiment No. 3, Table I The redistribution of salt was qualitatively similar to that which had been obtained in the absence of butyric acid. The amount passing into the filtrate was, however, smaller. Examination for Butyric Acid and Hemolytic Power.—Original: 48 per cent. butyric acid = (100) HBu. Heemolytic time = 8 minutes. Filtrates: The 1st and 2nd fractions contained (0°2) and (58) HBu respectively. The 1st fraction agglutinated the test blood strongly, and the 2nd hemolysed completely in 50 minutes. The 8th to the ilth fractions contained (90°6) HBu, and the hemolytic time was 12 minutes. Residual Fluid: Concentration (125) HBu. Hemolytic time, 5 minutes. Conclusions.—Butyric acid is retained by gelatine to a considerable extent. The gelatine appears to retain more sodium chloride in the presence of butyric acid. Liltration of Sodiwm Chloride through 9 per cent. Formalised Celatine. It seemed probable, from the marked change which takes place in gelatine on being exposed to formic aldehyde, that such gelatine would show a different permeability to that already found for ordinary gelatine. On the other hand, the possibility of investigating physiological fluids in an apparatus which could be thoroughly and easily sterilised by means of “formalin” might recommend the use of formalised gelatine filters. Method.—A filter containing 9 per cent. gelatine in 0°8 per cent. NaCl was fitted into the filter case, and the latter filled with a solution of 0°8 per cent NaCl containing 10 per cent. of commercial “ Formalin.” A few cubic centi- metres of fluid were pressed through, and the remainder allowed to stand in the filter case overnight. The solution was then removed, and, after about 12 hours, 100 ec. of 0°88 per cent. NaCl were placed in the filter case, and pressed through in fractions of 4 to 5 c.c. 1905. | Crystalloids and Colloids through Gelatine. 319 Lxamination for Sodium Chloride.—Experiment No. 1, Table I. Original Fiuid: 0°88 per cent. NaCl represented as Orig. (100). Filtrates: No formaldehyde or formic acid could be detected. From the 1st to the 11th fraction the NaCl gradually increased from (23:9) to (83°3), at 100 atmospheres filtration pressure. At this point the pressure was removed and filtration resumed after an interval of eight hours. Six fractions of 5 cc. each were then removed, and gave values decreasing from (112°6) to (93-7). Residual Fluid: Tested immediately after the last decompression con- tained (1349). Conclusions.—Formalised gelatine containing sodium chloride retains sodium chloride from a solution, on filtration, to a greater extent than ordinary gelatine. The decompression in this case also leads to the immediately following filtrates having a higher concentration of salt than the original fluid, but the effect is more marked than with ordinary gelatine. The concentration of the residual fluid is also much greater than with ordinary gelatine. Filtration af Megatherium Lysin Through Various Percentages of Gelatine. Megatherium lysin diffuses slowly compared with crystalloids in general,* and is probably of semi-colloidal character. It seemed probable that a substance of this type would be considerably more affected by a change in the concentration of gelatine than a good filter-passer such as sodium chloride. Hardy’s (1899) work on gelatine pointed also to the possibility of considerable change in the structure of the jelly at about 7 per cent. gelatine, which might lead to markedly different degrees of permeability between 7 and 15 per cent gelatine filters. Method.—Filters containing 15, 11, 9 and 7°5 per cent. gelatine in 0°8 per cent. NaCl were tested for tightness with 0°8 per cent. NaCl, of which 10 ¢.c. were pressed through in each case to ensure a concentration of NaCl in the succeeding filtrates of about 0°8 per cent. The fluid examined was the hemolytic filtrate from a broth culture of B. megatherium diluted with an equal volume of 0°8 per cent. NaCl. Filtration was carried through at a pressure of 100 atmospheres and a temperature of 10° to 12°C. and the filtrate was ‘collected in fractions of about 4 ce. Hxamination for Lysin—Table I. Original Fluid—The hemolytic index (see above) was (50). * Cf. Craw (IV, 1905). 320 Mr. J. A. Craw. On the Filtration of [Dee, 1, Filtrates: Experiments Nos. 4, 5,6 and 7, Table I, show the hemolytic values of the succeeding fractions through 15, 11, 9 and 7-5 per cent. gelatine respectively. The average indices for the whole filtrates were in the order given above (1°5), (6:9), (31°7), and (36°9). In Experiment No. 8, Table I, 50 cc. of 20 per cent. horse serum in 0:8 per cent. NaCl had previously been passed through a 75 per cent. gelatine filter. In this case the permeability to lysin was slightly less than that of a similar filter without serum, the average indices of ‘the filtrates being (30°6) and (369) respectively. Gelatine: On melting out the gelatine, after filtration, at 37° C., and mixing with the test blood, the latter was in all cases rapidly and completely heemolysed, index (100). Control experiments showed that saline on being pressed through gelatine which had been used to filter lysin became strongly hemolytic. The original gelatine had no hemolytic effect in the standard time. Residual Fluids: The residual fluids in all cases showed average indices. which were higher than that of the original fluid. The lowest portions of residual fluid had generally higher indices than the portions towards the upper surface. With rising concentration of gelatine the hemolytic power increased, thus with 7°5 per cent. the residue had an index (52°7) whereas. with 15 per cent. it was (85). Conclusions—Megatherium lysin is retained to a greater extent than sodium chloride, and more is retained with higher concentrations of gelatine. The residual fluids have higher concentrations than the original and the concentration is greater with higher percentages of gelatine. Diffusion and expression of lysin from the gelatine into the residual fluid are insufficient to account for the increased concentration immediately after decompression, and | it seems as if the water of the original fluid could pass into the gelatine more readily than the lysin. It seems possible, under the conditions of preparation, that during the draining of the filters from the various percentages of gelatine at 30° C., the filters with the higher concentrations would retain thicker surface layers of gelatine and so exaggerate the differences in permeability. Filtration of a Typical Colloid, Ferric Hydrate, through 11 per cent. Gelatine ‘ under Varying Pressure. As ferric hydrate in colloidal solution shows no appreciable tendency to diffuse, any redistribution of the colloid by the gelatine filter must be explained on some other basis than that of diffusion. Method.—A_ 5-per-cent. solution of colloidal ferric hydrate was prepared 1905. | Crystalloids and Colloids through G'elatine. 321 by adding ammonium carbonate to ferric chloride and dialysing for three weeks; it gave no trace of red coloration with potassium thiocyanate, but showed the characteristic reaction of colloidal ferric hydrate, viz., a slight yellowish precipitate. The filter used contained 11 per cent. gelatine which had been dialysed for 24 hours. The content of the fractions of filtrate, etc., in iron, was determined by converting the hydrate into chloride and estimating colorimetrically with potassium thiocyanate. Table II. Colloidal ferric hydrate. Neutral red. Iodine in potassium iodide. Rollin | Pres- Per- sail ae Pres- Per- Wola Pres- Per- Per- ne sure in | centage aie sure in |centage Ot sure in | centage} centage ; atmos. Fe. ne atmos. | N.R. “| atmos. I,. KI. 4 100 0:0 4 100 0:0 1 100 5 68 ‘8 f 4 100 0-7 4 100 2°3 4 100 8 69 °9 g 4 100 06 4 100 0:04] 4 100 17 70°9 £ 4 100 0-4 3 20 | 0:03] 4 100 | 19 69 9 P| 4 100 0-4 10 100 0°30) 4 100 21 69 ‘9 a 4 100 0-4 10 100 0-78 | 4 100 23 69 ‘9 S A 100 O 10 100 0:90] 4 100 25 68 ‘3 g 4 100 0:0 10 100 Waly |} al 15 20 69 ‘9 B 4 100 0-0 10 100 3°00 | 5 100 33 71°5 S 1 100 0:0 10 100 3:20 | 5 100 33 73°0 | eS 05 20 | 250-0 20 100 4°68 | 4 100 35 72:0 ry 1 100 4°75 = = = 0°75 (0) 10) 43 °O | a 2 100 0:9 — — — 0°75 | 100 25 93-5 | = 3 50 0-7 _ —- | = 6 100 37 90 °5 | an 05 15 66 6 — —- | — 6 100 50 86 ‘0 0°75 50 1:0 — — — il 0) 1 34 °4, 0:25 30 0% — — — 5 10 25 94 6 0°5 20 0°3 ~— — — 2 100 30 92 °5 0°75 10 O-1 — — — 3 50 30 96 °8 | 0°25 50 0-2 — — 4 100 30 91 °4 | 1 100 0:2 — — 5 50 30 | 85:0 5 15 0-4 — — — 5 25 30 | 82°8 4 100 0°4 = = = 5 25 25 | 79°5 0°5 15 0 “4 — — — 5 25 Py |) al | 4 100 0-4 | Res. | 50 == ))))/5180/-08. |) 20 — | 18-75 | 35 | eof teen | | | Orig.| 120 — 100 ‘0 120 — |100°0 {120 — | 100 | 100 ‘0 | | Examination for Ferric Hydrate-—Table II gives the relations obtained in one out of three similar experiments. Original Fluid: Five per cent. ferric hydrate = (100) Fe. Filtrates: The 1st and 2nd fractions contained (0:0) and (0°7) respectively, and in the succeeding fractions the amount diminished until in the 8th no 322 Mr. J. A. Craw. On the Filtration of [Dec. 1, trace could be found at 100 atmospheres filtration pressure. On suddenly diminishing the pressure to 20 atmospheres the liquid percolating through was very intensely coloured and gave the value (250). Re-establishing the pressure of 100 atmospheres, the filtrates became less and less intensely coloured (4:75) and (09). A drop in the pressure to 50 atmospheres did not materially change the concentration, but a further drop to 15 atmospheres gave a filtrate with the value (666). Increasing the pressure to 50 atmospheres caused the filter to tighten once more with respect to the hydrate, and on gradually decreasing the pressure to 10 atmospheres the filter remained tight. In this condition sudden variations of pressure from 100 to 15 atmospheres had but a slight effect on the permeability of the filter. Residual Fluid: The colour was more intense than that of the original fluid and its content was (180) Fe. Conclusions—The gelatine fifter is slightly permeable to the typical colloid ferric hydrate, but at constant pressure the permeability decreases as the filtration proceeds. The permeability is increased enormously by suddenly diminishing the pressure, but is not much affected by a gradual diminution. After gradual diminution of pressure a filter is obtained which does not markedly change in permeability on suddenly varying the pressure. The high concentration of the residual fluid is probably due to the water penetrating the gelatine easily, whereas the ferric hydrate is largely left on the surface of the gelatine, where it forms a concentrated solution of higher specific gravity than the rest of the fluid and so gives rise to convection currents which cause it partly to mix with the remainder of the fluid in the filter case. The candle retains a skin of colloidal ferric hydrate, and the filter most probably at the beginning of the filtration acted as a simple gelatine filter, but subsequently as a compound ferric hydrate gelatine filter. Filtration of a Staining Colloid, Neutrai Red, through 11 per cent. Gelatine under Varying Pressure. Preliminary experiments with horse serum and solubie starch showed qualitatively similar effects to those obtained with colloidal ferric hydrate, and it seemed probable that the majority of colloidal solutions would behave in a similar way on filtration through gelatine. It seemed probable, however, that those colloidal solutions which atia gelatine would show considerable difference in behaviour. Method.—A 0°5-per-cent. solution of neutral red in distilled water was pressed through 11 per cent. dialysed gelatine and the content of the fractions of filtrate, etc., determined colorimetrically. 1905. ] Crystalloids and Colloids through Gelatine. 323 EHzamination for Neutral Red.—Table II. Original Fluid: 0°5 per cent. neutral red = (100) N.R. Filtrates: The 1st, 2nd and 3rd fractions showed (0:0), (2°3) and (0:04) respectively. A sudden diminution in pressure from 100 to 20 atmospheres caused a slight decrease in the value of the filtrate (0°03). On raising the pressure once more to 100 atmospheres, the succeeding fractions gradually increased in content up to (4°68) N.R. , Residual Fluid: The colour intensity was much diminished and indicated only (18°75) N.R. Conclusions.—The gelatine filter is slightly permeable at the beginning of filtration to neutral red, but the permeability decreases, as in the case of colloidal ferric hydrate, on continued filtration at constant pressure. Diminution of pressure has the opposite effect to that obtained with sodium chloride and ferric hydrate as the permeability tends to decrease. The filter gradually becomes stained throughout and the permeability increases correspondingly. — Filtration of a Staining Crystalloid, Iodine in Potassium Iodide, through 11 per cent. Gelatine, under Varying Pressure. Iodine was chosen as a crystalloidal substance having the property of staining gelatine, and potassium iodide was used as its solvent in water. The interest of the experiment centred chiefly in the relative behaviour of the two substances and the mutual influence they exert on each other during filtration under varying pressure. Method.-—A solution of 0:214 per cent. iodine in 1°54 per cent. potassium iodide was filtered through 11 per cent. gelatine in 1°54 per cent. potassium iodide. The content of the fractions of filtrate, etc., was determined by titration with sodium thiosulphate, using starch as an indicator for the iodine and with silver nitrate, using potassium chromate as indicator for the potassium iodide. Examination for Iodine and Potassium Iodide.—Table II. Original Fluid: 0:214 per cent. Iz = (100) Iy. 1:54 per cent. KI = (100) KI. Filtrates: The 1st c.c. contained a considerable percentage of iodine (5) and likewise of potassium iodide (68°8). On continued filtration at constant pressure the iodine concentration rose steadily to (25); the iodide rose slightly and then fell to (68°3). On suddenly diminishing the pressure from 100 to 15 atmospheres the iodine concentration diminished and the iodide increased. On raising to 100 atmospheres the iodine rose, whereas the iodide rose and then fell in concentration. Removing the pressure for 12 hours the few drops of liquid which percolated through contained no iodine and only (43) of 324 Mr. J. A. Craw. On the Filtration oy [Dec. 1 iodide. Re-establishing the pressure of 100 atmospheres the iodine content rapidly increased, whereas the iodide in the first few drops was abnormally high (93°5), but decreased in the second fraction of 6 c.c. to (860). Once more removing the pressure for three hours the iodine diminished to (1:0) -and the iodide to (34'4), confirming the former result. The pressure was now slightly raised, viz., to 10 atmospheres, when the iodine became (25) and the iodide reached the highest value so far (946). A further rise in pressure to 100 atmospheres increased the iodine to (30) and diminished the iodide to (92°5). Variation between 100 and 50 atmospheres did not further affect the ‘iodine content but did influence the iodide. Finally,a gradual fall in pressure ito 25 atmospheres caused both iodine and iodide to diminish in concentration. Residual Fluid: The concentration of iodine was only (30), whereas the iodide was (109°7). Conelusions—The gelatine absorbs a large amount of iodine, but is easily permeable to the same and the permeability rises and falls, as in the case of meutral red, with rising and falling pressure. The increasing concentration of iodine in the gelatine increases the power of the latter in retaining potassium iodide. On entirely removing the applied pressure the gelatine robs the liquid percolating through almost entirely of its iodine and the iodised ‘gelatine abstracts about two-thirds of the iodide. The filtrates, after a drop in the pressure, show an increased amount of iodide on recompression similar to the effect observed in the filtration of -sodium chloride. On the Influence of the Nature of the Solution and of Varying Pressure on the Rate of Filtration. A few preliminary measurements of the rates of filtration of various -solutions were made to elucidate some of the physical properties governing the action of the gelatine during filtration. Method.-As a rough approximation the number of drops per minute falling from the nozzle of the filter case was taken as a measure of the rate .of filtration. The rate was determined after the first 2 c.c. had passed, and again after about 50 cc. in all had been collected. Rates of Filtration—Table III shows that distilled water, 0-8 per cent. sodium chloride and megatherium lysin pass 9 per cent. gelatine in the initial stage more rapidly than in the final under constant pressure. Solutions containing 0:214 per cent. iodine, with 1°54 per cent. potassium iodide, and 5 per cent. colloidal ferric hydrate pass 11 per cent. gelatine with similar ~variations in rate. For all these solutions the decrement in rate is practically 1905. | Crystalloids and Colloids through Gelatine. 325 the same. On filtering horse. serum in full strength, however, the decrement in rate was found to be abnormally large. Table III. 4 KI, | NaCl.| NaCl | NaCl NaCl | Dist. Coll. pepiaunios. 120 C-1) NaCl, I. | lysin. | toluol.| Am.Alc.| HBua. | water. | Fe(HO);. SST Initial drops per| 7 | wl 8 16 18 6 9 6 6 minute Final drops per 4 4, 5 18 60 18 4 4 1 minute A solution of 0°8 per cent. saline containing toluol as a fine emulsion did not show any decrement. On the other hand, 0°8 per cent. saline containing either amylic alcohol or butyric acid (48 per cent.) gave very distinct increments in the rate of filtration. Table IV gives a comparison of the rates obtained for distilled water and 0°81 per cent, sodium chloride containing 4°8 per cent. of butyric acid. It is evident that distilled water filters more and more slowly, whereas the saline butyric acid passes more and more rapidly. The effect of removing the pressure for two minutes enables the distilled water on re-establishing the pressure to pass through at a higher rate, and if the pressure be removed for 10 minutes the gelatine regains completely its original rate of filtration. Table IV. Distilled water. Sodium chloride and butyric acid. Pressure | Time in | Drops per| C.c. per | Pressure | Time in | Drops per | C.c. per in atmos. | minutes. | minute. minute. | in atmos. | minutes. minute. minute. 100 0 7 —_ 100 ¢) 6 = 100 13 6 0°5 100 2 5 = 100 21 5 0°5 100 5 9°5 = 0 23 (0) — 100 u 11 0-7 100 29 6°5 0°6 95 15 9°75 0-6 100 32 5 0°5 90 20 9 = 100 40 4:5 0°3 100 23 11 0-6 0 50 0 — 100 35 13 0-6 100 73 9 0-7 100 42 14 0-7 100 83 5°5 0°5 100 49 15 0-7 100 133 4°5 04 100 64 16 0°8 100 71 18 0:9 100 TA 18 aL Sit 326 Mr. J. A. Craw. On the Filtration of [ Dee. 1, Conclusion.—Solutions of salts such as sodium chloride, or colloids such as ferric hydrate, will not appreciably modify: the rate of filtration through gelatine unless in concentrated solution. Amylic alcohol and butyric acid accelerate the rate of filtration. It seemed possible that this effect might be due to the condensation of amylic alcohol and butyric acid on the walls of the gelatine pores, and that the change in viscosity might account for the increased rate. This, however, does not seem probable, for a toluene emulsion did not markedly affect the rate of filtration, and as the viscosities of the toluene and amylic alcohol used were found to be 0°49 and 4:4 respectively when compared with distilled water as unity, one might conclude that the toluene would accelerate, and the amylic alcohol retard, the rate of filtration. On the other hand, the action of amylic alcohol and butyric acid is consonant with the view that the surface forces between gelatine and saline are modified. The surface tension of saline against air diminishes with increasing concentration of both amylic alcohol and butyric acid. Both of these substances markedly influence the rates of sedimentation of solid suspensions through water,* and this fact seems to have a close connection with their influence on the rate at which water passes through a porous solid or jelly such as gelatine. The effect of continued pressure on gelatine seems to be a tightening of the pores, which is nearly complete at 100 atmospheres in 40 minutes, and decompression apparently allows the pores to resume their original dimen- sions in about 10 minutes at atmospheric pressure. On the Expression of Water and Gelatine from a Wet Filter. In agreement with E. W. Reid, the filtrates were found to contain gelatine, but in variable amounts. Thus the first fraction of about 4 c.c. usually con- tained most, and after 20 c.c. had passed the gelatine was only present in traces. Rough estimations with tannic acid indicated about 0°5 per cent. gelatine in the first, and less than 0:1 per cent. in the fifth fraction on filtering distilled water. These contents of gelatine were not found to influence the titrations given above. It has been hitherto assumed that the water in the pores of a wet filter is largely expressed before the liquid undergoing filtration appears in the filtrate undiluted by the same. For this reason Martin} discarded the first 10 c.c. of filtrate, and E. W. Reid} concluded that the concentration of even a second fraction of 25 c.c. might be considerably reduced by this dilution. It seems to me, however, that the effect of the water in the gelatine on the concen- * Cf. Craw (1904). + Loe. cit. t Loc. cit. 1905. | Crystalloids and Colloids through Gelatine. 327 tration of the filtrate is neither so marked nor so long continued as has been imagined, for the following reasons. In the first place, the filtrates from saline through gelatine containing the same concentration of saline are at first only about half as concentrated as the original saline. Secondly, on decom- pression, and again filtering, the filtrate may contain a higher concentration of salt than the original saline. This seems conclusively to show that much of the dilution of the filtrate is due not to the water of the filter, but to the retention of the salt by the gelatine. Further, Reid found the quantity of water in freshly prepared filters to vary between 2 and 13 grammes—~.c., pre- sumably, the amount which can be removed by dry air. It is probable that only a smal) part of this water will be removed by the passage of a slow current of liquid through the gelatine, the remainder being retained by adsorption or inbibition forces, etc. In this connection an observation which was made with all the filtrations given in this paper may not be unimportant. It was found that the sum of the volumes of the filtrate fractions and residual fluid in any one filtration experiment was less than the volume of the original fluid introduced into the wetted apparatus. It is a well-known fact that water can be forced into gelatine by. pressure, and as a large part of the gelatine in the pores of the filter, during filtration at 100 atmospheres, must be under considerable pressure, it seems reasonable to assume that part of the original water of the wet filter is more firmly bound, and that about 5 ¢.c. are imbibed partly from the water of the filter and partly from the fluid filtered. The rapid passage of iodine, neutral red, and colloidal ferric hydrate into the filtrate also point to a considerable percentage of the first fractions of filtrate being contributed by the fluid filtered. It seems, on the whole, as if the free water of the filter is almost wholly removed in the first fraction of 5 cc. filtrate. The Application of Mechanical and Adsorption Hypotheses to the Filtration Phenomena of Gelatine. Mechanical Hypotheses—The most obvious explanation of the retention of colloidal substances by the gelatine filter is that the colloidal molecule or grain is too large to pass through the pores. Martin* has advanced the view that the non-filtration may be due either to the size of the molecules or to some interaction between the colloidal membranes and the dissolved colloidal molecules. It has, however, been shown by E. W. Reid (1904), Gatin- Grazewska (1904), and others that proteids, glycogen, and other typically colloidal substances, exert no measurable osmotic pressure in solution, do not influence the freezing point of the aqueous medium, and in general diffuse * Loe. cit. VOL. LXXVII.—B. 2B 328 Mr. J. A. Craw. On the Filtration of [ Dee. 1, very slowly, if at all. These facts lend much probability to Oswald’s view that colloidal solutions may be regarded as very fine suspensions. As the molecular weights of these substances are as yet quite indefinite, it would seem more justifiable to speak of the colloidal “grain” than of the colloidal “molecule.” Accepting this view of the nature of colloidal solutions, the action of the filter admits of an apparently satisfactory explanation on a purely mechanical basis.* During the filtration, for example, of colloidal ferric hydrate, Table II, the first fraction contained none of the colloid, and probably consisted chiefly of the water of the filter. In the succeeding fractions a little ferric hydrate passed through the larger pores, but these were speedily blocked by the particles. Under the pressure of 100 atmospheres the pores are probably compressed, and the effective passages are soon com- pletely blocked by ferric hydrate. On suddenly diminishing the pressure the gelatine begins to return to its original condition, and the effective passages increase in number and diameter. The condensed ferric hydrate, no longer completely blocking the widened passages, is swept out of the gelatine by the oncoming fluid, and so gives a filtrate with an abnormally high concen- tration. On the other hand, by gradually diminishing the pressure the ferric hydrate has time to fill the new and the gradually-widening pores, so that the filtrate contains only a trace of colloid. If the ferric hydrate be not expressed on recompression, it is conceivable that the compound filter of ferric hydrate and gelatine would be less affected as regards permeability by variation of pressure than the original gelatine. Much as this view has to recommend it, it does not seem to be the whole explanation, as it does not satisfactorily account for the phenomena observed with sodium chloride. It is highly improbable that sodium chloride is retained by a mechanical blocking of the passages, and that it is swept in high concentration into the filtrate, on decompression, because the condensed sodium chloride is no longer of the same dimensions as the widened passages. Recourse must, therefore, be had to some other view, which may considerably modify the explanation given for colloidal substances. Adsorption Hypotheses—Porous substances, powders, and fine suspensions of solids in aqueous media have the power of removing salts and other substances from solution. The action is attributed to forces of the same nature as those which give rise to adhesion and the wetting of a solid by a liquid, eg., of clean glass by water. The phenomenon usually called adsorption is common to all substances with highly developed surfaces, and * Of. also M. Traube (1866, 1867). Traube regards precipitation membranes as atomic and molecular sieves. 1905. | Crystalloids and Colloids through Gelatine. 329 the intensity of the adsorption depends not only on the physical condition but also on the chemical nature of the active surface. Gelatine being, probably, a porous substance, presents a large surface to the enclosed fluid, and it seems highly probable that when this fluid contains crystalloids or colloids adsorption will take place. Thus, gelatine containing 0°8 per cent. saline apparently retains about one-half of its content in sodium chloride, and the retention of crystalloids and colloids on filtration through gelatine at constant pressure seems to admit of explanation on the basis of adsorption. The results obtained in the investigation of the adsorption phenomena of the sedimentation of silts through aqueous solutions may therefore be applied to what apparently is the converse of that process, viz., the passage of aqueous solutions through porous solids or powders, under pressure. This is rendered highly probable from the fact that the rate of sedimentation of a silt through saline is generally accelerated. by the presence of butyric acid, amylic alcohol, and other substances influencing surface tension, and similarly the rate of filtration through the porous gelatine is accelerated by the same substances. An additional factor may, however, come into play in gelatine filtration. The effect of pressure on gelatine containing a solution of salt will be to cause the absorption of water. It seems permissible to assume, in view of Hardy’s work on the structure of gelatine, that this additional water will be taken up by the web mass and will probably lead to a diminution in dimensions of the pores or web spaces. This view would account for the continued diminution in rate of filtration at constant pressure. The process should show some similarity to the passage of a solution into filter paper, where the water passes in more readily than the substance in solution. A sudden diminution in pressure will lead to the rapid expulsion of the imbibed fluid, which will sweep the adsorbed matter into the widening pores. The gelatine itself will thus express part of this concentrated fluid, and at low pressures highly concentrated filtrates will be obtained. This will take place markedly when the adsorption is easily reversible, but the less reversible adsorptions of certain staining substances, ¢.g., neutral red and iodine will not give this effect. Further compression and decompression would presumably lead to a more unitorm distribution of the adsorbed substance throughout the gelatine, and as the filter is now also more saturated, the effect of variation of pressure on the concentration of the filtrate would be less marked. Summary of Conclusions. 1. Wet gelatine filters are to be preferred to those which have been partially dried, as the former have more uniform rates of filtration, and dilu- tion of the filtrate by the water of the gelatine can be largely eliminated. 2B 2 330 Mr. J. A. Craw. On the Filtration of [ Dec. 1, 2. Under constant pressure the gelatine of the filter absorbs water, and its porosity gradually decreases ; on decompression this water is expressed, and the original porosity is rapidly regained. 3. Gelatine of a definite concentration apparently has a specific permea- bility for different crystalloids and colloids; the value is high but not complete for the crystalloids sodium chloride, potassium iodide, and butyric acid, and it is low, but not zero, for the colloids ferric hydrate, serum and soluble starch. 7 4. As filtration proceeds the crystalloids pass through in increasing con- centration, whereas the colloids rapidly decrease to zero. 5. The simultaneous filtration of two substances may influence their specific permeabilities, thus butyric acid alters the permeability to sodium chloride, and iodine that of potassinm iodide. 6. Variations in the gelatine influence the permeability, eg., formalised gelatine is less permeable to sodium chloride than ordinary gelatine, and 15 per cent. gelatine is less permeable to megatherium lysin than 75 per cent. 7. Variation in the pressure causes remarkable changes in permeability. A sudden diminution of pressure gives highly concentrated filtrates of both erystalloids and colloids, whereas a gradual diminution has practically no effect. 8. Substances which stain the gelatine, eg., neutral red and iodine, give filtrates with lower concentrations on diminishing the pressure. 9. The rate of filtration is accelerated by amylic alcohol and butyric acid, which accelerate the rate of sedimentation of silts in a similar way. 10. Part of the phenomena may be explicable on the mechanical view of a blocking of the gelatine pores, but chemical relations between the gelatine and substances filtered must be taken into consideration, and probably the most satisfactory view is that the action of gelatine on the solutions tested is essentially an adsorption phenomenon. 1905. | Crystalloids and Colloids through Gelatine. 331 REFERENCES. Arrhenius and Madsen (1902), ‘ Festskrift ved Indvielsen af Statens Seruminstitut,’ IIT. Arrhenius and Madsen (1904), ‘Bulletin de ?Académie Royale des Sciences de Dane- mark,’ p. 271. Bordet (1903), ‘ Annales de l'Institut Pasteur,’ vol. 17, p. 161. Brodie (1897), ‘Journ. of Pathology,’ p. 460. Brodie (1900), ‘ Brit. Med. Journ.,’ p. 300. Craw (1904), ‘ Lancet,’ p. 434. Craw (I, 1905), ‘ Journ. of Hygiene,’ vol. 5, p. 115. Craw (IV, 1905), ‘Roy. Soc. Proc.,’ vol. 76 B, p. 179. Ehrlich (1898), ‘ Deutsche Med. Wochenschrift,’ vol. 24, p. 597. Ehrlich (1903), ‘ Berliner Klin. Wochenschrift,’ vol. 40, pp. 793, 825, 848. Gatin-Grazewska (1904), ‘ Pfliiger’s Archiv,’ vol. 103, p. 281. Hardy (1899), ‘ Journ. of Physiology,’ vol. 24, p. 158. Krafft (1902), ‘ Zeitschr. f. physiolog. Chemie,’ vol. 35, pp. 364, 376. Landsteiner (1903), ‘Miinchener Med. Wochenschrift,’ vol. 50, p. 764. Martin, C. J. (1896), ‘Journ. of Physiology,’ vol. 20, p. 364. Martin and Cherry (1898), ‘ Roy. Soc. Proc.,’ vol. 63, p. 420. Nernst (1904), ‘ Zeitschrift fiir Hlektrochemie,’ vol. 10, p. 377. Reid, E. W. (1901), ‘ Journ. of Physiology,’ vol. 27, p. 161. Reid, E. W. (1904), ‘Journ. of Physiology,’ vol. 31, p. 438. Starling (1889), ‘ Journ. of Physiology,’ vol. 24, p. 317. Traube, M. (1866), ‘ Centralbl. f. d. Med. Wissenschaften,’ vol. 4, pp. 91, 113. Traube, M. (1867), ‘Archiv f. Anat. u. Physiologie,’ p. 87. 332 A Case of Regeneration in Polychete Worms. By Arnotp T. Watson, F.LS. (Communicated by C. S. Sherrington, F.R.S. Received October 23, 1905,—Read January 18, 1906.) The facts recorded in the following note were ascertained in the course of my observations made upon a marine rock-boring polychete worm found at Tenby in the spring of the present year. This worm, a species of Potamilla, is living in limestone rock, in which another species of the same genus, Potamilla reniformis, is, also burrowing. It differs from the latter in various particulars, amongst which may be noted the absence of eyes on the branchial filaments, the colour of the blood (which is red instead of being sometimes green), the form of the sete, and the character of the external tube, which is largely covered, especially at the tip, with minute pieces of shell attached edgewise, imparting to it a white, rugged appearance, somewhat similar te that of the tube of Owenia. The worm is sometimes as much as 33 inches long, and is seldom extracted entire from the rock, fragments only, of varying length, usually being obtained. It occurred to me that this material might be utilised for the study of the regeneration of the lost parts,and my experiments in that direction succeeded beyond my expectations. Not only did the fragments renew these parts (both anterior and posterior) but they demonstrated the existence of a power to economise labour in this respect, by changing the arrangement of certain of the old parts, so as to complete the model of the original animal. The body of the worm in question consists of a large number of segments, all of which, with the exception of those at the two extremities, are endowed with a set-of hooks (uncini) and bristles (sete) on either side, and it is one of the characteristics of the Sabellide, the family to which this worm belongs, that the character and arrangement of these appendages in the anterior or thoracic portion differs from that in the posterior or abdominal part. The sete in the former are situated dorsally and the uncini ventrally, whilst in the abdominal portion the uncini are dorsal and the sete ventral. This arrangement, besides enabling the worm to rotate on its long axis in either direction at will, also facilitates the bringing of its thoracic ventral glandular plates and collar-lobes into contact with the inner surface and top of the tube, and is probably connected with the tube-forming habits of the worm. Some of the fragments which I have had under observation were without head and thorax, and consisted of abdominal segments only, the sete and uncini consequently being ventral and dorsal respectively, from end to end. A Case of Regeneration in Polychete Worms. 333 In cases of regeneration my experience has been that new segments are much more freely. produced at the posterior than at the anterior extremity of the body, and the problem which occurred to me was, how, and in what length of time, the thoracic segments (about eight in number), with the inverted arrangement of cheetal appendages, which is needful to the life-work of the worm, would be reproduced. The answer came in the nature of a surprise. Beside the cephalic plume-bearing segment, one new setigerous thoracic segment only was formed, but the chetal plan of the succeeding five or nine abdominal segments was changed; the dorsal uncini in these segments first disappearing gave place to sete, and later the ventral setae were replaced by uncini; the new sets and uncini, moreover, were changed to the forms characteristic of this part of the body. In other words, so far as the chetal plan is concerned, a new thorax had been constructed from the abdominal segments. How far the internal structure has been affected by the change remains to be ascertained. The observations extended over a period of five weeks, and were made upon two portions of apparently one and the same worm. One portion (comprising the two parts marked @ and ¢ on the figure), about ? inch long and consisting of 69 abdominal segments, being minus head and thorax, as well as the anal, and numerous preanal segments; the other part, 0, } inch long (probably the hinder portion of c), consisting of 36 preanal and the anal segment. In order to expose a large surface of water to the air and bring the animals as near thereto as possible, the experiments were carried on in watch-glasses, fresh sea-water being supplied twice daily. The first fragment, ac, in course of a day or two, attached itself to the watch-glass by means of a narrow cord or loop which it secreted, and which served, by a constant twisting movement of the animal, to sever the portion, c, } inch long, and consisting of 18 of the posterior segments. The number of parts available for observation was thus increased to three. Each was placed in a separate watch-glass, to which it attached itself slightly by a secretion from the ventral surface, and each part succeeded in changing the chetal arrangement of certain abdominal segments into that of thoracic ones, as follows—in a, nine segments became thoracic ; in ©, five ; and in 0, nine. The normal number of thoracic segments in the few adult specimens which have passed through my hands is eight (in one case it is nine), but the number appears to vary considerably in any yviven species of Sabellid. The number of abdominal segments transformed is possibly, to some extent, regulated by the total number of segments contained in the fragment under- going repair, but it may also be dependent upon the number of thoracic 334 Mr. A. T. Watson. [Oct. 23, mS SALA A a » Z aD poe») DS =D, —. aa == =n = ED c =D =D =D == { ==> — feos See aaa ax, booo bee) jaca =p ss) com & =) rm’ =D oa = = =D ==zD es) ). The cells exhibit an increase in cytoplasm, a comparative absence of secretory activity, and a peculiar and well-defined change in the appearance of the nucleolus. The alteration in this last-named structure consists in its larger size and denser appearance. Furthermore, very many nuclei are to be seen in a state of active division. Whilst some exhibit various stages of mitosis, others are clearly undergoing fragmentation or amitosis. At this stage of the development of the tumour, the peripheral cells that are dividing mitotically show all the characters of ordinary premaiotic divisions, and the normal number (32) of chromosomes can frequently be counted with certainty (figs. 5 and 10). But concomitantly with the first changes indicated in the epithelial cells at the edge of the neoplasm, a marked activity may be observed to take place on the part of the leucocytes. These bodies are seen to be in a condition of active migration and multiplication, much like that which occurs during the early stages of simple inflammation. In the subsequent stages, however, the early parallelism with inflammatory processes is lost, and there supervenes a remarkable phase in the further development of the cancerous cells. Not only do the cells of the tissue in question multiply with great rapidity, whilst the leucocytes amongst them are enormously increased in number, but the latter are seen not infrequently to force their way into the cancer cells, particularly in the so-called “giant cells,’ where, however, they are still to be recognised with ease and certainty (figs. 1,2, d,11,12,13). This circumstance has already been noticed by others, but we have been led to attach a somewhat special importance to its occurrence. Some writers have suggested that the cancer cells are acting phagocytically upon the leucocytes, but, as a matter of fact, the further sequence of events indicates that the cancer cells are no more to be regarded as attacking the leucocytes than the latter as destroying the cancer cells. There can be no possible doubt that the leucocytes actively force their way into the elements in question. They may not seldom be observed to be in close juxtaposition with these, or in a hollowed depression, or finally they may be discovered just within the cell membrane, where they are easily recognised on account of their characteristic nuclei (fig. 11). They show no sign of disintegration—at least, in the great majority of cases—and the fact that they may persist for a considerable time without destroying the cell into which they have invaded, is proved by examples in which a leucocyte lying in the cancer cell is seen to be surrounded by several nuclei that have clearly originated by the frag- mentation of the original cell nucleus, and, indeed, one of these is shown to be still dividing amitotically. 1905. | On the Cytology of Malignant Growths. 341 But the strongest proof of the persistence of the leucocyte under these remarkable conditions is afforded by the cases, not few in number, in which we have been able to trace the leucocyte actually dividing within the cancer cell (figs. 12,13). Of course, it is only during the early stages that it is possible to be certain that a second dividing nucleus in a mass of protoplasm belongs to a leucocyte, and does not represent mitosis in a small nucleus that has arisen by fragmentation. But we have seen so many cases of early stages of leuco- eytic mitosis within the cancerous (or “precancerous”) cell that it seems impossible to resist the inference that many of those frequently occurring cases in which a small nucleus is seen in the later phases of mitosis within the large nucleated cancer cell are to be attributed to this source. The nuclei of the cancer cell and leucocyte often divide simultaneously, and the two nuclear figures may also coalesce more or less intimately, and thus a commingling of leucocytic and epithelial chromosomes occurs on a spindle that becomes common to the two nuclei concerned. The cells so affected were, as already stated, usually the very large (giant) cells so characteristic at this stage of the development of the tumour, and we found that more than one leucocyte might enter and persist in a single cancer cell. In the earlier stages, of course, there is no difficulty in clearly recognising the intruding cell, since it retains its own cytoplasm and lmiting membrane intact (see fig. 11), and the highly characteristic structure of the nucleus enables it to be identified even after these criteria have ceased to exist. In the same region in which this series of events is proceeding a number of cancer cells are to be seen in various phases of mitosis, and, both in the aster and diaster of such nuclei, larger numbers of chromosomes were often encountered than are proper to normal somatic cells. These increased numbers are partly to be ascribed to the pluripolar mitosis distinguished by Hertwig and by Von Hansemann, and they result from the simultaneous mitosis of a number of nuclei lying in a common cytoplasmic mass. But the observations recorded above indicate that, in the addition of leucocytic nuclei to those of the actual epithelial cells, we have confronted, at any rate, with one of the sources to which these excessive numbers of chromosomes (hyperchromatic nuclei of Von Hansemann) may be attributed, although a large number of the cells continue to multiply in the manner already described, it may also be seen that there exists a very considerable amount of amitosis, or direct nuclear divisions in the cells of the young parts of the tumour. There appears to be no evidence which would point to the conclusion that amitosis is in any way bound up with degeneration, or diminishing activity in those cells in which it occurs. Elements that have previously multiplied by amitosis and by fragmentation have given rise to 342 Prof, Farmer, Messrs. Moore and Walker. [Nov. 17, the highly characteristic multinucleate cells, may again assume the mitotic method of increase, and vice versd. A curious feature in the further division of these multinucleate cells, or syncytia as they may, perhaps, be more appropriately termed, is seen in the almost invariable circumstance that, on the resumption of mitotic activity, all the nuclei are in exactly the same phase. This simultaneous character of the process is one which is shared by many other syncytia, eg., the myxomycetes. In these organisms, the nuclei are commonly observed not only to be dividing simultaneously over a considerable area of the plasmodium, but they also exhibit identical phases of the process at any given time. In examples of this simultaneous mitosis within the neoplastic syncytia, it often happens that the spindles of some, or even all of the dividing nuclei, become more or less intimately fused together, and in this way various forms of pluripolar mitosis are produced. Probably these pluripolar divisions owe their origin chiefly to the cause just indicated. The figures produced are extremely variable, and it not unfrequently happens that, whilst the chromosomes belonging to the different nuclei are ageregated in the centre, the poles of three or more of the spindles involved are quite separate. In other examples the groups of chromosomes do not coalesce, but each equatorial plate is quite distinct, and lies in a plane different from that occupied by the equatorial plates of the other spindles. But when a more intimate fusion of the ends of two or more spindles takes place, it is obvious that the daughter nucleus formed in relation to such unions will receive an excessive number of chromosomes. We would call special attention to the fact that giant cells of this character, also containing several nuclei, are present not only in the normal human testis, but also in the so-called red bone marrow, and that pluripolar mitosis may occur in such cells in a manner precisely similar to that so characteristic of cancerous tissue. The divisions of these early cancerous cells also exhibit other characters likewise encountered in the cells of the testis. Very often the daughter chromosomes do not move regularly towards the poles, but some either stray out of the direct line, or in other ways occupy unusual positions. These figures are also well known to occur in the heterotype division of some spore mother-cells of plants. In yet other examples of divisions in cancerous tissues, we have confirmed the observation of Von Hansemann that some of the chromosomes, as they are passing to the spindle poles, get ahead of their fellows, and form isolated or grouped chromatic particles that look as if they are about to be left out in the cytoplasm when the daughter nuclei become reconstituted. These figures 1905. | On the Cytology of Malignant Growths. 343 are also paralleled by similar occurrences that may be seen in the cells of the testis,and they are known to occur during the maiotic divisions of some plants, It is thus evident that hyperchromatic nuclei, that is, nuclei containing an excessive number of chromosomes, may be produced in at least two ways: firstly, by the inclusion of leucocytes, and the incorporation of the chromo- somes belonging to these bodies with those of the cancer cells when mitosis sets in; secondly, through the formation, whether by amitosis or mitosis, of multinucleate syncytia, and by the subsequent confusion and mixing of the chromosomes originally belonging to two or more of the nuclei when the equatorial plate stage is reached. These aberrant modes of division are found to proceed concurrently with the normal somatic mitoses that are going on in other cells in their immediate vicinity. It is impossible to say definitely whether there may exist any sort of alternation between the two types, though we are inclined to think that such is not the case. It is, however, important to notice that all the mitoses described above, whether they are normal in the number of chromosomes or not, agree in conforming to the somatic type of division. That is to say, no matter how many or how few the number of chromosomes involved may be, the spireme eventually divides into a number of rod-like elements, each of which splits longitudinally, and the daughter chromosomes resulting from such fusion are severally distributed between the daughter nuclei finally produced. In such typical cases this of course means that each of the two daughter nuclei receives one longitudinal moiety of such original chromosome. But as we pass inwards from the growing edge of the tumour we encounter cells in which the nuclei exhibit important deviations from the ordinary somatic type of mitosis, and exhibit the characters otherwise met with during the heterotype division (¢/. figs. 6, 7, 8). In the early stage of the phase of such nuclei the spireme exhibits that characteristic bunched appearance recalling the well-known contraction figure that is normally to be seen at the onset of the maiotic phase, that is in the prophase of the heterotype mitosis, in animals and plants. In addition to this, we have been able to ascertain that at about the same stage the spireme thread exhibits the longitudinal fission (fig. 6) that is highly characteristic, though perhaps not exclusively confined to the prophase of the heterotype division. The fission is especially well seen in those cases in which a marked polarisation of the spireme is apparent. But the most striking evidence of the validity of the comparison that we drew in 1903 between these particular nuclei and those of the reproductive cells during the maiotic phase of the animals and plants does not depend solely on the similar mode of evolution of the chromosomes from the resting nuclei in the “gametoid” cancerous and the true reproductive VOL, LXXVII.—B, 2¢ 344 Prof. Farmer, Messrs. Moore and Walker. [Nov. 17, elements. The number of the chromosomes furnishes a far more important criterion. It is seen that a large number of dividing nuclei contains less than the normal complement of chromosomes. We have made a number of careful counts of the chromosomes in numerous cases of carcinoma, and always with the same result. In especial, we are indebted to Mr. L. Robinson for his assistance in this somewhat trying task. He has estimated the chromosomes in 400 dividing nuclei, taken (100 from each) near the actively growing regions of three different carcinomata originating respectively from the rectum, scrotum, penis, and in an example of deciduoma malignum. In every case we find two well-defined maxima, one set of nuclei containing 32, the other 16, chromosomes. For purposes of comparison he has counted chromosomes of the testis of the cockroach (Curve F), so as to obtain a control indicating the probable degree of accuracy represented by the estimations in the cancer nuclei. The same two maxima are, of course, apparent, but there is a similar average error around the maxima, due to the difficulty of the actual counting, and also the chance that some of the chromosomes might be absent from the section, or that a limited degree of variation may really occur. And, having regard to the fact that in the human species the chromosomes are not easy, even under favourable conditions, to estimate very accurately, whereas in the case of the cockroach the observer encounters far less difficulty in this respect, the results may, we think, be described as satisfactory. For although, after what we have said, it is obvious that, owing to amitosis, and especially to pluripolar mitosis, a considerable extent of variations is to be anticipated, the grouping of the numbers around the maxima of 32 (somatic) and 16 (reduced) is quite unmistakeable, as 1s shown in the accompanying curves. ie ee | Number of cells Carcinoma of Rectum. The ordinates represent the number of cells that contained any given number of chromosomes, as indicated by the abscissz. A. Cancer of the Rectum.—The hypochromatic nuclei to the left somewhat obscure, the maximum at 16. The grouping of numbers about 24 and 64 are fairly well shown. 1905. | On the Cytology of Malignant Growths. 345 in lad (i, anni Number of ch eae 2 | Number of cells Epithelioma of the Scrotum. Number of cells CI 10 15 20 25 30 35 40 45 50 55 60 Number of chromosomes Epithelioma of the Penis. a te” AlN tit | IS Io I5 i 40 Number oF Eirencseen 5 un Number of cells ° — Deciduoma malignum. B. Epithelioma of the Scrotum. The maxima about 16, 24, and 32 are distinct, that about 64 not so clear. C. Epithelioma of the Penis.—The maxima in the regions of 16, 24, 32, 48, and 64 are all fairly distinct. D. Deciduoma malignum.—There is considerable irregularity in the nuclei in this growth, which was somewhat advanced, and deviations are therefore to be anticipated. Clee Namommeielyeemasciwl emp b | Mec ee fede | Nest ANA HN | | LAM AANA ee DectoohluNh wNueer oF dn Bieosance E. Combined Curve drawn from the Results shown in the preceding Four Cases of Cancer, viz., carcinoma of rectum, epithelioma of scrotum and penis, and deciduoma malignum, The three maxima about 16, 32, and 64 are unmistakeable. DG D 4 25 20) 15 am fe) eh Number of cells. 346 Prof. Farmer, Messrs. Moore and Walker. [Niowaraltyis 30 25 pce} Number of cells w 8e is 20 25 30 35 40 umber of chromosomes z z Testis of Periplaneta Americana. F, Curve obtained by countings made from dividing nuclei of the maiotic and premaiotic cells of the testis, in order to estimate the probable error in the far more difficult cases of cancer. It will be seen that there is some not inconsiderable variation about the two maxima of 16 and 32. This is due partly to underestimating the number of chromosomes actually present, and partly to the nuclei having in some cases been partly damaged in preparing the section. We shall further consider this matter in the concluding part of the paper: at present we are mainly concerned with showing that there exists a striking resemblance between what we have termed the “gametoid” cells of cancer and the cells of normal reproductive tissues, and as we pass to the later phases of mitosis we find the same loop and barrel-shaped chromosomes present in both, and we have occasionally seen, during the diaster of a cancer nucleus, the late longitudinal fission in the daughter chromosomes as they diverge from each other, just as it occurs in the heterotype diaster of so many animals and plants. An inspection of the curves shows the relative frequency of the different numbers of chromosomes met with in the younger cancerous areas. Whilst, as already pointed out, the two maxima of 16 and 32 are unmistakeable, it is also obvious that amongst the generally irregular numbers two other groups occur with greater frequency than others. Thus, there is a distinct indication that nuclei containing about 24 chromo- somes may be regarded as forming a distinct group, also that a second, though far less well-marked, series is characterised by containing about 64 (double the normal somatic number) chromosomes. It may be that the latter are related to the ingression of the leucocyte already described, but it is difficult at present to guess at the significance of the grouping of 24. There is no obvious indication that the nuclei with 48 chromosomes are specially common, and, in the absence of more direct evidence, it is useless to indulge in speculations that may prove to be devoid of all foundations. 1905. | On the Cytology of Malignant Growths. 347 In animals, as has already been stated, it invariably happens that, after the onset of the first maiotic (heterotype) mitosis, there ensues only one further nuclear division, commonly designated as the homotype, on account of its close general resemblance with a normal somatic mitosis. The principal point of constant difference lies in the retention in the former of the reduced number of chromosomes. The cells originating from this division give rise after a more or less complex series of changes of form and of the inter-relation of their constituent parts and the sexual cells without any further intervening nuclear divisions. In plants this is not the case. The cells issuing from the homotype mitosis always undergo one or (often) many subsequent divisions beforé some or all of the resulting units develop into sexual cells. It is therefore of interest to find in cancerous tissue that there is abundant evidence that the cells, the nuclei of which have undergone reduction, are capable of continued division, and, indeed, a great part of the tissue of the cancer is made up of such cells, which, in accordance with the terminology we have elsewhere employed, we may term post-maiotic, or “gametoid.” It will be seen that we differ from Von Hansemann in our explanation of these “hypochromatic” nuclei, regarding them as have arisen, not as the author just named believes, by a dropping-out of chromosomes from the spindle, or through some form of degeneration, but chiefly as the result of a process resembling, or identical with, that by which reduction is ordinarily effected in the tissues destined to give rise to the gametic cells. But we desire to definitely state that, in using the term “gametoid,” we expressly differentiate between the cancerous cells and those of normal reproductive tissues. The relation existing between them, if any, is at present obscure ; and, though we think the resemblances, which will be still further emphasised by facts we are about to describe, are very suggestive, we are far from holding the views which have been expressly or implicitly ascribed to us by other writers as to the identity of gametic with “ gametoid” cells and tissues. Finally, then, it is clear that there exist in the facts of pluripolar mitosis, on the one hand, and in amitosis on the other, a mechanism sufficient to explain all the irregular numbers encountered in a young cancer. But the irregularities, while masking, cannot conceal the far more frequently recurring numbers of chromosomes, whereby the reduced (halved) and, though far less frequently, the double, numbers become apparent. But the existence of the irregularities indicated above often renders extremely difficult the task of deciding to what category a particular departure from the normal somatic number is to be relegated. There is a further body of evidence bearing on the resemblance between’ 348 Prof. Farmer, Messrs. Moore and Walker. [Nov. 17, cancerous and normal reproductive tissue to be derived from a study of the so-called Plimmer’s bodies of cancer.* It was shown by one of us in 1895f that, during the prophase of the first maiotic (heterotype) division of the spermatogenetic cells of mammals, the archoplasm undergoes a peculiar and definite series of metamorphoses. In ordinary somatic or premaiotic cells, this body is seen to le beside the nucleus as a dusky mass of protoplasm, in the centre of which are found the centro- somes. Thus, in these cells, the attraction sphere consists of the archoplasm plus the centrosomes (fig. 3, 0, fig. 4, a). When, however, we turn from the premaiotic or the somatic cells to the prophase of the heterotype (first maiotic) mitosis, we find these’ two constituents have become separated (fig. 4, b). The centrosomes migrate from the centre of the archoplasm, and are eventually seen to le outside that body, and*completely detached from it (fig. 4, ¢). At the same time the archoplasm itself undergoes a change, small vesicles are developed in its substance (fig. 14), and, at the close of this particular cell-generation, both vesicles and archo- plasm become merged and lost in the general cytoplasm of the daughter cells. In the prophase of the second maiotic (homotype) mitosis the same peculiar phenomena recur, and the archoplasm and the vesicles, in lke manner, become lost during the later stages of this (homotype) division. In the spermatids, which result from it, the persistent centrosomes can be readily seen to be perfectly disconnected with the new archoplasm which is differentiated in these cells. The archoplasm becomes filled with minute vesicles, as in the two preceding cases, subsequently the vesicles enlarge, and they either fuse together, as in some mammals, or one usually takes the lead and grows larger than the rest, as commonly happens in the guinea-pig and in man (fig. 15). The body thus formed was originally termed the archoplasmic vesicle in 1895,t and it is a very conspicuous and constant feature peculiar to the sperm cells of the vertebrata, whilst it has also been encountered by various observers in animals outside that group. When fully developed, the archoplasmic vesicle often assumes a size approximating to that of the nucleus itself, the latter being often deformed into a crescentic shape, owing to the enlargement of the vesicle that hes adjacent to it in the cell. In normal spermatids, the vesicle and its contents ultimately form the so-called “cephalic cap” of the spermatozoon (fig. 16, @). * See ‘Roy. Soc. Proc., vol. 76 B, “On the Resemblances existing between the ‘Plimmer’s Bodies’ of Malignant Growths and certain Normal Constituents of Repro- ductive Cells of Animals,” by J. Bretland Farmer, J. E. 8. Moore, and C. E. Walker. + Moore, ‘ Internat. Monatschr. f. Anat. u. Physiol.,’ 1894. {t Moore, loc. cit. 1905. | On the Cytology of Malignant Growths. 349 Now, the “Plimmer’s bodies” are well known in the cells of many cancerous growths (fig. 17), and they are most commonly met with in the young erowing portions of the tumour. They appear in the form of vesicles, and consist eventually of a fairly well-defined wall, enclosing a clear space, in which is suspended a smal] and densely refracting granule. They appear to occur with greater frequency in cancers of a glandular or glandular-epithelial origin.* They lie in the cytoplasm of the cancer cells, usually in close proximity to the nucleus. They vary im size from excessivery minute bodies to forms as large as the nucleus itself. The special interest attaching to the Plimmer’s bodies depends on the fact that they have commonly been regarded as peculiar to cancer cells, although Hondat believes that he has occasionally encountered them in inflammatory tissue. They have, in fact, been variously interpreted. Some investigators have regarded them as parasitic organisms, more or less intimately connected with the etiology of the disease, whilst others have seen in them a differentiation of the cancerous cell itself. Borrelf suggested that they might represent hypertrophied centrosomes, but the observations of Benda,§ who showed that centrosomes and Plimmer’s bodies coexisted in the same cell, have rightly been held to disprove the view advanced by Borrel. When the foregoing facts are all taken into consideration, the case originally upheld by ourselves|| appears to be a strong one. We see no escape from the position that the Plimmer’s bodies of cancer represent the archoplasmic vesicles that occur in the normal reproductive cells at the stages already indicated. And this forms an important link in the chain of similarities connecting cancerous tissue with the normal reproductive elements. But in this relation it is of interest to note that we have recently observed bodies, which appear to be closely similar to archoplasmic vesicles, to occur at apparently definite stages in the life history of certain leucocytes which are present in bone marrow. General Conclusions. To sum up the observations already recorded in this paper, it may be seen :— * Greenough, ‘Third Report of the Caroline Brewer Croft Cancer Com.,’ Hary. Med. School, 1905. + Honda, ‘ Virch. Arch.,’ vol. 174. t Borrel, ‘Ann. Inst. Past.,’ vol. 15. § Benda, ‘ Verh. deutsch. Gesellsch. f. Chir.,’ 1902. || ‘Roy. Soc. Proc.,’ vol. 76 B, pp. 230 et seq. 350 Prof. Farmer, Messrs. Moore and Walker. [Nov. 17, 1, That a primary growth originates in the first instance as the result of a change in the nature of a number of previously functional somatic cells. 2. That the transformation may affect a considerable number of cells, and certainly continues to operate for some time. 3. That, as the result of the change, mitotic and amitotic activity is awakened, and proceeds rapidly, with a consequent increase in the mass of affected tissue. 4. That during this increase a remarkable activity prevails amongst the - leucocytes, at first resembling that seen in inflammatory processes, but finally leading to the union of at any rate some of the affected cells with one or more leucocytes. 5. That in the subsequent divisions of these cells the nucleus of the leucocyte divides simultaneously with that of the cancer cell, and their chromosomes may become mingled in cleavage figure. 6. That multinucleate cells (syncytia) may arise by mitosis or by amitosis, unaccompanied with the division of the mass of protoplasm. 7. That the resulting nuclei may divide normally and mitotically, or the nuclear figures may be more or less mingled, and hence all sorts of variations in the number of chromosomes may occur. But the mode of chromosome evolution and division follows the somatic type. 8. In addition, a form of mitosis occurs, leading to nuclei with half the number of somatic chromosomes, and the phases closely accord with those observed during the heterotype (first maiotic) mitosis of animals and plants. 9. Subsequent divisions occur, in which the reduced number of chromo- somes is retained, the type of division otherwise resembling that of ordinary somatic cells. These mitoses fall into the category corresponding with the post-maiotic mitoses of plants. 10. During the maiotic and post-maiotic divisions in the cancerous cells, structures are present which have been designated as Plimmer’s bodies. These are common to cancerous cells and to the reproductive cells of the testis at a particular phase in their evolution. The only other cells in which structures resembling the bodies in question have been observed are possibly those forming certain of the leucocytes in bone-marrow. It will be evident from the above summary that the change from the healthy to a cancerous development is intimately bound up with definite change in the cells affected. The onset of the change is probably to be attributed to the operation either of new stimuli upon the body cells, or to a change in the constitution of the latter. Such an alteration might originate in a variety of ways. For example, it might be ascribed to the influence of a parasite. But we have never succeeded in tracing any such cause, and it 1905. | On the Cytology of Malignant Growths. 351 becomes necessary therefore to seek for some other explanation for the phenomena actually witnessed. It is quite certain, in the first place, that we are dealing with the trans- formation of functional somatic cells into cancerous ones, and this, to our own minds, affords a complete refutation of the hypothesis as to the persistence of “embryonic rests,” such as have been supposed by Cohnheim and his followers to account for the incidence of the disease. We have drawn attention to the events that occur in connection with the invasion of the cells of the young growths by leucocytes, and, although we are fully aware that further investigations into the details of these processes are required before a final opinion can be expressed as to their true significance, the facts themselves are very suggestive. Furthermore, the interest attaching to these fusions is not lessened by a study of the bone-marrow, in which the leucocytes can be most advantageously observed. For we have seen in this tissue all the abnormal types of nuclear and cellular division that are so highly characteristic of cancerous cells, and we have ascertained a fact of even greater importance, namely, that some of the nuclei of dividing marrow cells certainly possess less than the full complement (52) of somatic chromosomes. We would, further, lay emphasis on the occurrence, in the same preparations of bone-marrow, of other cells in which the process of mitosis was strictly somatic in character, both as regards the form and number of the chromosomes. But it is none the less certain that the other nuclei exhibit chromosomes of a remarkable form, elongated in the direction of the spindle, and strongly resembling those which are so characteristic of the heterotype mitoses of the testis or of a cancer. Whilst it is obvious that further investigation on the cytology of bone-marrow is urgently needed, it is evident that, if it should ultimately prove that the cells which are derived from the results of fusion of a leucocyte with a tumour cell really represent the progenitors of the malignant elements themselves, a satisfactory explanation would be afforded not only of the striking nuclear character of the diseased tissues, but also of the invasive and destructive powers they undoubtedly possess. The destructive action of the leucocytes themselves on other cells of the body, especially during old age, is too well known, owing especially to the valuable researches of Metschnikoff, to call for further comment here. Such a view of the case as is here tentatively suggested is not in conflict with the idea embodied in the term “gametoid” tissue, but rather forms an extension of it. We have, as already pointed out, from the first maintained the existence of a resemblance, extending to extraordinarily minute detail, between the “ gametoid,” cancerous, and the reproductive tissue, 352 Prof. Farmer, Messrs. Moore and Walker. [Nov. 17, which, in the case of animals, gives rise to the gametes immediately after maiosis. But it is also now certain that there exist certain striking similarities between the leucocytic and reproductive cells which are, in themselves, highly suggestive, and this is not diminished by a consideration of the earlier phylogenetic history of wandering and reproductive cells in more primitive animals, for example, in sponges. For the present, however, and in the absence of more complete and accurate knowledge on the evolution of the leucocytes, we may close by remarking that the various peculiar characteristics of cancerous cells find thei closest analogies in the cytological processes that are exhibited in the formation of the reproductive cells, and in those maiotic phenomena that so especially distinguish them. DESCRIPTION OF PLATES. PLATE 8. Fig. 1.—Section of the growing edge of a young Carcinoma of the Rectum. lw, 1y,1z. Enlarged parts of the same drawing. The letters c, d,e, correspond with those on the main figure. a. The portion to the right represents the normal structure of the rectum ; 6 the zone in which transmutation from healthy to cancerous tissue is proceeding. c. Cell showing somatic division (see also 1 z). d. Cells in this zone containing leucocytes (see also 1 z and 17). e. Cell showing prophase of first maiotic (heterotype) mitosis (see also 1 2). f. Cut portion of crypts, but belong to the zone of transformation. g. Portions of the growth invading the adjacent layers. PLATE 9. Fig. 2.—Section through young Epithelioma of the Penis. 2x, 2y, 22. Enlarged parts of the same drawing. The letters a, b, c, d, correspond with those of the main figure. a. Cells showing somatic (premaiotic) divisions. b. Cells showing somatic division, but with excessive number of chromosomes. c. Cell showing first maiotic (heterotype) division. d. Cell with leucocyte in its cytoplasm. - Puate 16. Fig. 3.—Small Portion of the Testis of a Guinea-pig, showing (a) premaiotic cell dividing ; (b) ceil in prophase of first maiotic (heterotype) division. In this it will be seen that the centrosomes are at the centre of the archoplasm. Fig. 4.—Portion of the Testis of a Guinea-pig, showing (a) cells with the synaptic contrac- tion, and the normal condition of the attraction sphere ; (6) late stage in the prophase of the first maiotic division, showing the centrosomes detached from the archoplasm ; (c) homotype prophases showing. The same dismembered condition of the attraction spheres. Fig. 5.—Cell from the early Cancer of the Rectum given in fig. 1, showing the somatic character of division. Compare with fig. 3, a. Fig. 6.—Cell from Cancer of Rectum given in fig. 1, showing the characters of the prophase of the heterotype division. Compare with fig. 4, a. * Farmer, Moore and Walker. Roy. Soc. Proc., B. vol. 77, Plate ues il. i he a 2ae Ny Feadeguol (Biber Ove Oy © 595° - y 7 ue Lf) YO W541 Ba ee, lite, 2. Farmer, Moore and Walker. voy. Soc. Proc., B. vol. 77, Plate 10. Na fates: Farmer, Moore and Watker. Roy. Soc. Proc., B. vol. 77, Plate 11. Soc. Proc., B. vol. 77, Plate 12. Roy. Farmer, Moore and Walker. Fig. 4. 1905. | On the Cytology of Malignant Growths. 353 Fig. PuatTe 11. 7.—Cell from an example of Decidua malignum, showing the later phases of the heterotype mitosis. . 8.—Similar Cell from an Epithelioma of the Tongue. . 9.—Cell from the Testis of Man, showing the later stages of the heterotype division. Compare with figs. 7 and 8. . 10.—Cell from a Cancer of the Rectum, showing the somatic or premaiotic character of the chromosomes and the large number of these elements. . 11.—Cell from an early Cancer of Rectum, showing the peculiar condition of the nucleus, which suggests amitosis ; also two leucocytes (a) within the cytoplasm. . 12.—Cell from the same showing nucleus in the prophase of division, and also an intruded leucocyte (@), with its nucleus in the same phase. . 13.—Cell from Cancer of the Rectum, showing nucleus in division, and that of intruded leucocyte (@) in a late prophase. PLatE 32. . 14.—Portion of the Testis of a Guinea-pig, showing spermatids with developing archoplasmic vesicles and centrosomes. . 15.—Portion of the Testis of a Guinea-pig, showing a later stage in the development of the archoplasmic vesicle. In this the origin of the tail of the spermatozoon is also seen, in connection with one of the centrosomes. . 16.—Portion of the Testis of a Guinea-pig, showing the remains of the archoplasmic vesicle becoming converted into the so-called “cephalic cap” (a) of the spermatozoon. . 17.—Cells from a Cancer of the Breast, showing Plimmer’s bodies and the position of the centrosomes. Compare with figs. 13 and 14. 304 On the Sexuality and Development of the Ascocarp of Humaria granulata Quél. By Vernon H. Buiackman, M.A., Assistant, Department of Botany, British Museum; Late Fellow of St. John’s College, Cambridge ; and HELEN C. I. Fraser, B.Sc., Assistant Lecturer, Royal Holloway College. (Communicated by Professor Marshall Ward, F.R.S. Received October 31,—Read December 14, 1905.) [Puates 138—15.] The observations of Harper(15, 16, 17) on Spherotheca, Erysiphe and Pyronema, have clearly shown that some at least of the Ascomycetes exhibit an ordinary sexual process. It is true that attempts have been made by Dangeard (10) to refute Harper’s observations, and doubt has been cast on his work by Lindau (22), Holtermann (18), and others; but the recent very convincing work of Claussen (8A) on Boudiera,* together with the strong circumstantial evidence obtained by Barker(1 and 2} in Monascus and Ryparobius, by Miss Dale(9) in Gymnoascus, and by Baur(3, 4), and Darbishire (104) in lichens, and, also, the confirmation by ourselves (7) of Harper’s work on Spherotheca, leave no doubt that the sexuality of the Ascomycetes is founded on a firm basis. The earlier non-cytological observations of a number of forms, however, . have shown clearly that the existence of a normal sexual process can hardly be expected in all the Ascomycetes. For example, in Melanospora parasitica Kihlmann (20) observed the development of the archicarp into the perithecium without the intervention of an antheridium; in Chetomium, Oltmanns (24) found that the antheridium was usually absent; in Ascobolus, the earlier observations of Woronin (25) and Janczewski (10), and the later observations of Harper(16), brought to light no definite antheridium. Again, among the lichens, according to the observations of Fiinfstuck (13), in Peltigera and Peltidea the ascogonia are without trichogynes ; and Solorina saccata, according to Baur’s (4) researches, seems clearly to develop without any ordinary fertilisation. It is obviously, then, very desirable that the cytology of some member of the Ascomycetes, the ascocarp of which develops without fertilisation by an antheridium, should be carefully investigated. The form here studied is of this type and hence is of peculiar interest. Humaria granulata Quél (= Peziza granulata Bull), a common Dis- * The form investigated by Claussen would seem to be more correctly placed in the genus Ascodesmis (vide Fr. Cavara, ‘ Annales Mycologici,’ vol. 3, 1905, p. 363). Sexuality, etc., of the Ascocarp of Humaria granulata. 355 comycete about 5 mm. in diameter, and of a yellow, orange or reddish tint, is found growing on the dung of various animals, especially of cow, and is most abundant during autumn and winter. The spores, apparently, normally germinate only after they have passed through the alimentary canal of the animal, for artificial cultures could not be obtained. A preliminary peptic or tryptic digestion or a combination of both seemed to have no effect on germination. Only a small number of experiments were made in this direction, for by bringing the material into the laboratory, natural cultures can sometimes be obtained, in which the fungus occurs in such abundance in appropriate stages that the necessity for artificial cultures is completely obviated. The material was chiefly fixed in Fleming’s weak fluid, which was allowed — to act either for 24 hours or for one hour, fixation being completed in the latter case with Merkel’s fluid. Either safranin, gentian violet and orange, or Benda’s iron-hematoxylin were used for staining. The very youngest stages of the apothecia are of course quite invisible, even with a powerful hand-lens, but sections of them were secured by removing and fixing the superficial layers of the substratum on which young apothecia were just visible. The behaviour of the closely-packed nuclei of the ascogonium was best followed in sections 4 w in thickness. Vegetative myceliwm.—The vegetative mycelium consists of cells which show numerous nuclei, but these, unlike those of the ascogonium, are not at all well marked, but appear generally as slightly staining homogeneous or granular bodies which sometimes show a minute nucleolar dot (Plate 13, fig. 1). The cells of the whole vegetative mycelium and of the apothecium contain a number of fairly large spherical granules, which stain deep red with the safranin of the triple stain. Im all the hyphe of the vegetative mycelium and many of those of the ascocarp, these granules are found collected in eroups on opposite sides of the transverse walls (figs. 1 and 4). These groups of granules were observed by Woronin(25) in this form and in Ascobolus, and by Harper (17) in Pyronema, but their function is unknown: Harper has suggested that they may have something to do with the passage of material from cell to cell through the wall. No reproductive organs other than the apothecia were observed in connection with this form. Development of Apothecium. As long ago as 1866, Woronin (25) showed that the apothecium began by the development of an archicarp as a side branch of an ordinary hypha of the mycelium. He observed that the apical cell of this branch was round and very much swollen, and that, later, side branches grew up from the cells 356 Mr. Blackman and Miss Fraser. Sexuality and [Oct. 31, of the stalk and completely invested the apical cell. The large cell he considered to be of the nature of an egg, while one of the branches growing up from below he thought was probably of the nature of an antheridium:; he was unable, however, to follow any further details of development. Actual Observations.—As Woronin observed, the first beginning of the archicarp consists of a branch with a variable number of somewhat short cells (figs. 2A, 2B, 3). The apical cell of this row is the ascogoniwm* which soon increases in size and becomes spherical, and exhibits beautiful vacuolate structure (figs. 24, 2B). The lower cells also increase in size and both they and the ascogonium become closely filled with food-material, so that the whole archicarp has a dense and opaque appearance (figs. 4, 5). Before the ascogonium has attained its full size a number of narrow branches begin to erow out from the cells of the stalk immediately below (fig. 4). These are the first beginnings of the investing hyphe which soon grow up and com- pletely cover in with layers of plectenchyma the ascogonium and upper few cells of the stalk (figs. 5,9). In cleared preparations, however, the cells of the ascogonium and of the stalk, owing to their greater density, can, for a time, be distinctly seen threugh the investing sheath (fig. 5). The number of cells in the stalk is variable ; there may be only a few, as is apparently the case in fig. 3, but usually a large number are to be observed, as in fiy. 5. From one of the cells about the middle of the stalk a small side branch was sometimes seen which grew down into the substratum and apparently aided the stalk in absorbing nourishment. None of the hyphe which grow up from the stalk act im the way Woronin suggested; they are all mere vegetative investing hyphx, no antheridiwm being developed. The ascogoniuw shows a vacuolate protoplasm with a number of nuclei which are better defined than those of the vegetative cells; with the growth of this organ these nuclei become much more distinct, exhibiting a nuclear membrane and a single deeply-staining nucleolus (figs. 6, 7), but no chromatin is to be observed in the nuclear cavity. As development proceeds the nuclei increase only slightly in size but enormously in number ; and the small vacuoles are replaced by one or more large ones (figs. 8, 10). At about the stage when the vegetative hyphe completely surround the ascogonium, the wall of the latter becomes thickened and shows a distinct differentiation into two layers, the outer, thin and deeply staining, the inner, thicker and lightly staining (figs. 10, 11). ' The wall between the ascogonium and the uppermost stalk cell exhibits at a young stage the usual apposed groups of granules, but at a later stage the * For a discussion of this use of the terms archicarp and ascogonium, vide infra. 1905.| Development of the Ascocarp of Humaria granulata, 357 granules apparently fuse together, for when the ascogonium has reached its full size this wall shows two large and deeply-staining masses placed opposite one another on either side of the wall (fig. 13). The masses sometimes show a central deeply-staining portion, and an outer, irregular, less dense portion (fig. 13). When the ascogonium and stalk cells become emptied these masses disappear. Besides the special accumulation on the walls a number of large granules are usually to be found scattered in the cytoplasm of the ascogonium and stalk cells (fig. 9). When the ascogonium has become covered in with several layers of vege- tative hyphe the ascogenous hyphe appear as narrow, thin-walled outgrowths from the thick-walled ascogonium, and make their way through the close mass of investing hyphe (figs. 10, 11). Into the ascogenous hyphe there pass nuclei and cytoplasm from the ascogonium, which becomes more and more vacuolate in appearance and is finally almost completely depleted. | It is clear that the ascogonium which produces the ascogenous hyphe has undergone no process of fertilization by male nuclei, so the development at first sight appeared to be a truly parthenogenetic one. When, however, such a case as the development of the accidium of Phragmidiwm violaceum (Blackman, 5) was considered—where, in the absence of the male cell, there is a peculiar process of fertilization by the union of a vegetative cell with the female cell—it seemed conceivable that a reduced process of a somewhat similar nature might be found in H. granulata also. If this were so, two possi- bilities presented themselves ; either the ascogonium might be fertilized by the entrance of the contents of the uppermost stalk cell or of some other vege- tative cell, or a fusion in pairs of the nuclei of the ascogonium might take place. As no evidence of the first possibility could be obtained, the ascogonial contents were very closely examined at various stages of development, with the result that the second hypothesis was found to be correct, and the female nuclei were observed fusing in pairs in the ascogonvum. These fusions are to be observed in ascogonia of various ages, sometimes when the investment of the ascogonium has only just begun, but usually at some stage between investment and the emptying of the ascogonium. There thus appears to be no definite stage of fusion for all the nuclei corresponding to that of Pyronema, but a gradual fusion in pairs takes place as development of the ascogenous hyphe proceeds. The majority of fusions were observed when the ascogonium was partly emptied of its contents, as the nuclei are then not so deeply crowded as in earlier stages, and the cytoplasm does not stain so deeply. The nuclei in most stages are so close together that it is usually impossible to distinguish from mere accidental contact the contact of nuclei which is a 358 Mr. Blackman and Miss Fraser. Sexuality and [Oct. 31, preliminary to fusion. In a number of cases, however, nuclei were found in pairs more or less isolated from their fellows, which were probably to be considered as on the point of fusion. The actual fusion of nuclei seems to take place very quickly, for such a stage as that of fig. 14 is rarely seen, but the nucleoli apparently remain for some time separate, for the nucleus with two nucleoli (figs. 15, 16, 17) is found more frequently. Apart from the stages of contact, which must necessarily be impossible to distinguish with certainty, more than 11 cases of actual fusion were observed, so there can be no doubt that the fusion of the ascogonial nuclei in pairs is a regular process. The size of the nuclei is not of much help in deciding whether a given nucleus is or is not the result of fusion, as three or four different sizes of nuclei may be observed in a single ascogonium (fig. 18). The nuclei apparently undergo a fairly rapid growth in size, those at the centre of the ascogonium being usually smaller, at least in the later stages, than those at the periphery. The number of nuclei in the ascogonium varies apparently with the size of that structure, but in order to gain some idea of the number in an average ascogonium the nuclei were counted in two cases in a series of sections of an ascogonium. In the young ascogonium, of which a section is shown in fig. 7, the number 336 was determined, while the older one of fig. 10 gave 824. These countings are, of course, only approximate, as the nuclei are very crowded, and lie sometimes one above the other; also in the older asco- gonium, a small number of the nuclei had already migrated, and some of those still remaining had, no doubt, already fused. The number produced by the division of the original nuclei of the ascogonium might therefore be taken as about 1000. No data were obtained as tu the number of nuclei in the ascogonium at its first inception, but judging from the size of the organ at that stage and from the relatively small number of nuclei in the vegetative cells, very numerous divisions must take place. It is curious that such divisions were never observed in the ascogonium ; it 1s probable that they are intermittent in occurrence; possibly they take place only at night. When the nuclei pass out into the ascogenous hyphe they show a very distinct nucleolus and are easily defined structures (fig. 11); thus the distine- tion of the ascogenous from the vegetative hyphe (fig. 19) among which they ramify is rendered possible. The ascogonium becomes finally emptied of its contents, though sometimes a few nuclei and a little cytoplasm remain behind at the periphery. Soon after the ascogonium becomes empty, the connection of the ascogenous hyphe with it becomes obliterated, and these hyphz appear as independent struc- tures. This result is no doubt brought about by the pressure of the 1905.] Development of the Ascocarp of Humaria granulata. 359 surrounding cells which leads to a slight collapse of the wall, aud so to the obliteration of the cavity of the ascogenous hyphe at their point of origin. Ultimately the whole ascogonium becomes obliterated, though it remains con- spicuous as a large empty vesicle up to the time when the ascocarp first opens (fig. 31). The stalk cells also become emptied (fig. 30), and are obliterated somewhat earlier, so that after the early stages of opening no trace at all of the archicarp can be observed. It may be mentioned here that the nuclei of the uppermost stalk cell are generally more distinct than those of the ordinary vegetative hyphe, being intermediate in structure between those and the ascogonial nuclei (figs. 11 and 12). In the earlier stages of development the whole of the nourishment for the growth of the apothecium is supplied by the archicarp from its reserve of material, the cells of the stalk supplying the branches which arise upon them, while the ascogonium supplies the ascogenous hyphe. In the later stages, however, a “secondary mycelium” is formed consisting of vegetative hyphe which grow down into the substratum and absorb nourishment which is Supplied to the vegetative hyphe of the ascocarp, and so indirectly to the ascogenous hyphe which, after the emptying of the ascogonium, are practically parasitic on the vegetative hyphe. The first asci are formed very early before the outer peridium is burst through (fig. 30); they arise on the ends of the ascogenous hyphe by the peculiar process of the bending over of the apex and the fusion of the nuclei in the sub-terminal cell (figs. 20 to 25), such as has been described by Harper, Dangeard, Guillermond, and Claussen. In two cases the ascus was observed in a terminal position as described by Maire(23) and by Guillermond (14) for Galactinia succosa. When the two nuclei have fused in the ascus, the fusion nucleus begins to increase in size and to show a definite chromatin substance between nucleolus and wall. The division of the nucleus does not call for any particular comment, as it is not very favourable for investigation ; the spindles are at first intranuclear and show well-marked centrosomes with radiations, but the chromosomes are too small to allow of an estimation of their number (figs. 26 to 28). The method of spore formation in the ascus appears to be of the well-known type first described by Harper, but owing to the density of the contents of the ascus and the somewhat small size of the spores the object is not a favourable one for the study of the details of the process (fig. 29). The paraphyses at their first appearance form a wedge-shaped mass, which appears to play a part in bursting open the peridium, as described by Harper in Ascobolus. VOL. LXXVIIL—B. 2D 360 Mr. Blackman and Miss Fraser. Sexuality and [Oct. 31, The structure of the mature apothecium is of the ordinary type; there isa definite parenchymatous peridium, a well-marked .hypothecium consisting of large cells; and the paraphyses are large and club-shaped and filled with orange granules (fig. 31). General Considerations. It is clear that the process of fusion in pairs of the female nuclei in the ascogonium of Humaria granulata must be considered as a reduced sexual process which, in the absence of the antheridium, replaces the normal fertilisa- tion by male nuclei such as we find in Spherotheca, Erysiphe, Pyronema, and Boudiera. It renders even more untenable the most recent view of Dangeard* (10) that in the Ascomycetes as a whole there is no fertilisation in the ascogonium, but the sexual process has been shifted from that structure to the asci; for in HA. granulata we find that even in the absence of the antheridium the process of nuclear fusion is not confined to the asci, but there is an earlier fusion in the ascogonium, which must itself be considered as the sexual process, although of a reduced type. As stated earlier, the question of the occurrence of an ordinary sexual process in some at least of the Ascomycetes must now be considered as com- pletely settled. Future work must decide how far the members of the group exhibit ordinary sexuality or the reduced process described above; it is possible also that some forms are truly parthenogenetic,+ while there appears to be no doubt that others, as the Hxoascacee, are still further reduced, the asci having a direct vegetative origin. _ It can hardly be denied that the process of fusion of the female nuclei in pairs is derived by reduction from the ordinary sexual process such as we find in Pyronema; therefore it seems best to class such a process as a “yveduced sexual process ” (Blackman (5)), in which the male gamete has been replaced by another female cell (nucleus), the ecidium, just as in Phragmidium violaceum (5) the male cell is replaced by a vegetative cell. * Kuyper (21) in a recent paper, published since these observations were complete, has come to a conclusion somewhat similar to that of Dangeard. He has investigated Monascus and finds there only a single nuclear fusion, and that in the ascogonium, but without relation to the male nuclei. He considers Monascus a primitive form and that in the other Ascomycetes the fusion has been shifted to the ascus. Different results have been obtained by other workers on Monascus, and Kuyper’s figures are not very con- vincing ; but if there is only a single fusion, such a fusion is obviously comparable to the first fusion in Humaria granulata and not to the second. + That is with potential female gametes developing without any process of cell or nuclear fusion. If there be a true alternation of cytologically distinct generations in the Ascomycetes this is not likely to occur, as true and complete parthenogenesis is unknown in plants possessing such an alternation. 1905.| Development of the Ascocarp of Humaria granulata. 361 Davis (12) has criticised such a terminology in the case of Phragmidiwm, and objects to the use of the terms fertilization or sexual process being applied to any union in which the fusion is not between the regular male and female cells. He would class these irregular processes under the head of asexual fusions. It is true that a fusion in which the special sexual cells do not both take part cannot, from a purely morphological point of view, be a sexual process. When, however, it is considered that in some of these irregular fusions one of the sexual cells actually takes part, and also that they are of very special nature, being directly related in the phylogeny of the group to the ordinary sexual process, in fact, replacing that process in the life history, they can hardly be satisfactorily relegated to a class of asexual unions, where they are herded with processes most of which have not been shown to have any connection with true sexual fusions.* If on strict morphological grounds these fusions are separated from true sexual processes they should obviously be made a class apart, quite distinct from the asexual unions.t Itis doubtful, however, whether a purely morphological test of a sexual process (syngamy) is desirable when we consider that the process is essentially a physiological one and that primitively it occurs between vegetative cells (¢.g., Spirogyra, some Protozoa). Further, these irregular processes show no characters for which a parallel cannot be found in other accepted sexual processes ; for in the simplest cases the fusing cells are not differentiated, and in other cases of sexuality the blood-relationship between the fusing cells (eg., lateral conjugation in Spirogyra, sexuality in Basidiobolus and many Phycomycetes) is apparently as close as in the process under discussion. Since, then, these special processes in themselves have no characters which remove them from the class of sexual unions, and since they take place * Such as the fusion of nuclei in endosperm cells, and in cells which have been placed under abnormal conditions, the fusion of nuclei in the ascus, the “ vegetative” cell fusions in the Floridez, ete. + In the present state of our knowledge the cell and nuclear unions among plants would seem to be best divided into four classes :— (1) Cell-unions of an ordinary sexual nature : (2) Reduced sexual processes as described above : (3) Nuclear unions, such as are found in the teleutospore, basidium, and perhaps those of the spores of the Ustilagine ; these (at least in the case of some Uredinee and probably in the other cases) are the direct result of sexual, or reduced sexual, processes which exhibit nuclear association without nuclear fusion : (4) Asexual cell and nuclear unions, which are of doubtful or purely vegetative nature. The third class is of very special nature, and it is not satisfactory to class them, as does Davis, with the asexual unions. BAD We 362 Mr. Blackman and Miss Fraser. Sexuality and [Oct. 31, at a definite point in the life-cycle, and replace in phylogeny the ordinary sexual process, it seems proper that their relations should be exhibited in the terminology, and they should be classed as sexual processes or fertiliza- tions, with the addition of the term “reduced,” which indicates that one, or both, of the regular sexual cells has been replaced by some other cell. In the case of Phragmidium violacewm* and the “apogamous” prothallia, we may consider that there has been a sudden return in part, or as a whole, to the primitive condition where every vegetative (gametophytic) cell is a gamete. It appears, then, from the study of H. granulata that the female ceenogamete possesses a very striking property—the capacity to fertilize itself. It may be that in this capacity lies the explanation of the development without male sexual organs which seems normal for a large number of Ascomycetes ; future research can alone settle this point. Although the sexual process to be observed in H. granulata is, of course, morphologically reduced in relation to the normal sexual process, yet physiologically there can be little to choose between the fusion of ascogonial nuclei, which may be separated in descent by many divisions, and the ordinary sexual fusion in which, as is often the case, the antheridum and ascogonium are intimately related in origin. In fact the kinship of the fusing nuclei may very likely be closer in Boudiera, where sexual organs are borne in pairs on the same hypha and contain a small number of nuclei, than in Humaria, where the number of nuclei in the ascogonium is very great.f As has been suggested elsewhere (Blackman (6)), the majority of the (morphologically) normal sexual fusions in the Fungi, exhibiting as they do close-related sexual organs, are already physiologically reduced in relation to the typical (and probably primitive) exogamic sexual process. The morphological reduction found in the special fusions is thus only a small step which does not affect the physiological nature. In other words, instead of the fusion of the gametes from two gametangia borne close together and in intimate relation on the same plant, we have the abortion of the one and the fusion in pairs of the gametes of the other; put in this * The case of the xcidium of another species, P. speciosum, in which neighbouring cells fuse in pairs, described by Christman (8), and considered by him as a simple process of conjugation of undifferentiated gametes, would seem to be much better interpreted as a reduced sexual process, in which, in the absence of the male cells (spermatia), the female gametes fuse in pairs, as in 4. granulata (vide 7A). + It is not asserted that the close kinship or otherwise of the fusing nuclei necessarily makes any physiological difference, but that judged by this standard the processes are essentially similar ; and it is not clear that there are any other physiological factors which would differentiate the two processes. 1905.| Development of the Ascocarp of Humaria granulata. 363 way it is clear that there is a morphological difference, but a physiological difference is not easily conceivable. When one considers the apparent physiological equivalence of the ordinary and the reduced sexual processes, the ease with which “self- fertilization ” can be carried on in the ccenogamete, the small number of forms in which an ordinary sexual process has been observed, and the fairly large number which appear to have no antheridium, it seems not improbable that the reduced sexual process will prove to be the more common type of fertilization in the Ascomycetes. It is obvious that the occurrence of fusions among the nuclei of the female ccenogamete itself renders still more difficult the investigation of the sexual cell of this type. The mere presence of a male organ and the observations of nuclear fusions in the female cell is now not sufficient to prove a normal fertilization ; nor even is continuity between male and female organs, for the male nuclei may degenerate in situ, and a reduced fertilization of the H. granulata type may take place. To prove the existence of ordinary fertilization, evidence must be obtained for an actual migration of male elements to the female organ. Jt might perhaps be suggested by some that the nuclear fusion observed by Harper in Pyronema were really fusions between female nuclei like those in H. granulata. MHarper’s observations on the passage of the male nuclei into the oogonium seem, however, sufficiently satisfactory to allow of this > supposition being put on one side. Dangeard’s observations on Pyronema are very probably to be explained by the supposition that he was working on a form with a functionless antheridium. He worked with artificial cultures, while Harper used natural ones, and it has been shown by Van Tieghem (26), in a paper which seems to have been overlooked in the discussion, that Pyronema is very susceptible to artificial conditions. In his cultures Van Tieghem observed forms which were normal, forms which showed the ascogonium and antheridium reduced in size, and lastly forms in which the antheridium was absent, but the ascogonium developed normally. Dangeard was probably investigating a form in which the antheridium, though still present, had already become function- less ; in the light of the series of forms observed by Van Tieghem, one cannot conclude with Dangeard that the antheridium is always functionless. In such a case as Pyronema with a functionless antheridium a “ reduced fertilization” similar to H. granulata is to be expected; such a process would almost certainly be overlooked unless attention was specially directed to it. The other forms lately investigated by Dangeard (10), in which either the antheridium was absent or the male nuclei degenerated, may, perhaps, 364 Mr. Blackman and Miss Fraser. Sexuality and ([Oct. 31, also be explained by the fact that a reduced fertilization in the ascogonium was overlooked. Exact data as to the nuclear behaviour of Ascobolus furfuraceus in its early stages of ascocarp development will be of special interest. It would seem likely that the nuclei fuse in pairs when they meet, as described by Harper (16), in the large ascogonial cell which gives origin to the ascogenous hyphe. One of Harper’s figures shows these nuclei in very close contact. It is possible, also, that the parthenogenesis in other forms, in which a coenogamete develops without the intervention of a male organ (eg., the Mucorini and Saprolegniacez), may be explained in the same way, by a fusion, in pairs, of the female nuclei,* and so not be a true parthenogenesis. The fusion in the ascus still remains a most puzzling process, for which, at present, no explanation is forthcoming. That it is not a substitute for the ordinary sexual process, nor a nuclear fusion which has been shifted, in descent, from the ascogonium to the ascus, as Dangeard and Kuyper believe, is clearly shown (apart from such forms as Pyronema, Boudiera, etc.) by H. granulata, where, even in the absence of the antheridium, the fusion in the ascus is preceded by a fusion in the ascogonium. On the other hand, the curious simultaneous division of the two nuclei at the time of ascus formation—whether the ascogenous hypha bends over at the apex or whether it remains straight, as in Galactinia succosa—seems only to be explained as a method of ensuring that the fusing nuclei are separated in descent by at least one division. Now, such a separation, in descent, of the fusing nuclei is, as far as we know, an attribute of sexual fusions alone (though in many fusions which are accepted as sexual, the degree of relationship is very close). We have thus two closely related fusions, one of which is obviously a sexual fusion, while the other, in one character at least, partakes of a sexual nature. A satisfactory solution of the difficulty of the dual fusions can hardly be expected till we know the number of chromosomes throughout the life-history of some ascomycete. That there is a definite alternation of generations in the life-history of Ascomycetes which possess an ascogonium seems very probable. The ordinary vegetative mycelium would appear to be the gametophyte, which bears the ascogonium, and antheridium if present, while the products of fertilization, the ascogenous hyphe (which are parasitic on the gametophyte) and the asci, represent the sporophyte. The countings of chromosomes are, however, too few and too unsatisfactory to allow of * Kuyper (21) has independently made a similar suggestion in the case of the Saprolegniacez, and has even suggested that the figures of Davis (11) on egg development in Saprolegnia support the view of a nuclear fusion. 1905.| Development of the Ascocarp of Humaria granulata. 365 a decision. as to the cytological distinction of the two generations. It would seem, also, that there must be two reductions, as there are two fusions. The three divisions in the ascus might be expected to show at least one reduction, but Harper(17) is of the opinion that the number of chromosomes remains unaltered during these divisions. Of course, it is possible that the second fusion is of a peculiar nature and does not lead to a doubling of the chromosomes. What is obviously necessary is the dis- covery of a sexual ascomycete with a small number of distinct chromosomes in its nucleus, so that the number can be observed throughout the life- history.* De Bary uses the term archicarp as practically synonymous with ascogonium. It seems much more satisfactory to use the term archicarp for the whole fertile branch, apart from the antheridium, and to confine the term ascogonium to that part of the archicarp the contents of which take part in the formation of ascogenous hyphe, 7. the reproductive cell or cells which contain the female nuclei. It is in this sense that the terms have been used in the body of the paper. Used in this sense the term ascogonium is not necessarily confined to the cell or cells actually giving origin to the ascogenous hyphe. In Ascobolus furfuraceus, for example, the whole curved fertile branch, or scolecite, is the archicarp; the central part would be the ascogonium, which is divided into a number of cells by a series of perforate septa, as Harper (16) has shown; only one of the cells of the ascogonium, however, actually gives origin to the ascogenous hyphe, though the contents of all the ascogonial cells pass into this special cell and so into the ascogenous hyphe. In Melanospora parasitica, from the observations of Kihlmann, the ascogonium is represented by one or two cells of the archicarp, though the cytological details are not known. In Pyronema, Humaria, Spherotheca, and Erysiphe the ascogonium is a single cell, and naturally gives origin to one or more ascogenous hyphe. In Collema the archicarp consists of a few small sterile cells at the base, then comes the ascogonium, which is multicellular, and above is the multicellular trichogyne ; all the cells of the ascogonium appear to give origin to ascogenous hyphe. * The view that there are two reduction processes is also put forward by Harper in a very important paper (“Sexual Reproduction and the Organization of the Nucleus in certain Mildews,”) received while the present paper was passing through the press. 366 Mr. Blackman and Miss Fraser. Sexuality and [Oct. 31, LIST OF PAPERS. Barker, B. T. P. “The Morphology and Development of the Ascocarp in Monascus,” ‘ Annals of Botany,’ vol. 17, 1903, p. 167. Barker, B. T. P. “Further Observations on the Ascocarp of Ryparobius,” Leaflet, Brit. Assoc. Meeting, 1904. Baur, E. “ Die Anlage und Entwickelung einiger Flechtenapothecien,” ‘ Flora,’ vol. 88, 1901, p. 319. Baur, E. “Untersuchungen iiber die Entwickelungsgeschichte der Flechten- apothecien,” ‘ Bot. Zeit.,” vol. 62, 1904, p. 21. Blackman, V. H. “On the Fertilization, Alternation of Generations and General Cytology of the Uredinez,” ‘ Annals of Botany,’ vol. 18, 1904, p. 323. Blackman, V. H. “On the Relation between Fertilization, ‘Apogamy’ and ‘ Parthenogenesis,’” ‘ New Phytologist, vol. 3, 1904, p. 149. Blackman, V. H., and Fraser, H.C. I. “ Fertilization in Sphzrotheca,” ‘ Annals of Botany,’ vol. 19, 1905, p. 567. . Blackman, V. H., and Fraser, H.C.1I. “Further Studies on the Sexuality of the Uredinez” (zbzd., vol. 20, 1906, pt. 1). Christman, A. H. “Sexual Reproduction in the Rusts,” ‘Botan. Gazette,’ vol. 39, 1905, p. 267. . Claussen, P. “Zur Entwickelungsgeschichte der Ascomyceten. Boudiera,” ‘ Bot. Zeit.,’ vol. 63, 1905, p. 1. Dale, E. “Observations on the Gymnoascaceze,” ‘Annals of Botany,’ vol. 17, 1903, p- 571. Dangeard, P. A. “Recherches sur le Développement du Périthéce chez les Ascomycetes,” ‘Le Botaniste,’ 9e série, 1904, p. 59. . Darbishire,O. V. “ Die Apothecienentwickelung von Physcia pulverulenta,” * Jahrb. f. wiss. Botanik,’ vol. 34, 1908, p. 329. Davis, B. M. “ Oogenesis in Saprolegnia,” ‘ Botan. Gazette,’ vol. 35, 1903, p. 233. Davis, B. M. “Studies on the Plant Cell,” ‘The American Naturalist, vol. 35, 1905, p. 217. Fiinfstuck, M. “ Beitrige zur Entwickelungsgeschichte der Lichenen,” ‘Jahrb. d. K. Botan. Gart. zu Berlin,’ vol. 3, 1884, p. 155. Guilliermond, A. ‘Contrib. 4 étude de la Formation des Asques,” ‘Rev. Gén. de Botanique,’ vol. 16, 1904, pp. 49 and 129. Harper, R. A. “Die Entwickelung des Peritheciums bei Spherotheca Castagnei,” ‘Ber. d. d. Bot. Geo.,’ vol. 13, 1895, p. 475. Harper, R. A. ‘Ueber das Verhalten der Kerne bei der Fruchtentwickelung einiger Ascomyceten,” ‘ Jahrb. f. wiss. Botanik,’ vol. 29, 1596, p. 655. Harper, R. A. “Sexual Reproduction in Pyronema confluens, etc.,” ‘Annals of Botany,’ vol. 14, 1900, p. 321. ‘Holtermann, ©. “Mykologische Untersuchungen aus den Tropen,” Berlin, 1898. Janezewski, E. ‘Morphologie des Ascobolus furfuraceus,” ‘ Botan. Zeit.,’ vol. 29 1871, p. 257. Kihlmann, O. “Zur Entwickelungsgeschichte der Ascomyceten,” ‘ Acta Soc. Se. Fennice,’ vol. 14, 1885, p. 309. Kuyper, H. P. “Die Perithecien-Entwicklung von Monascus purpureus, Went. usw.,” ‘ Annales Mycologici,’ vol. 3, 1905, p. 32. Lindau, G. “ Die natiirlichen Pflanzenfamilien,” Teil 1, Abt. 1, 1897, p. 323. Maire, R. “Recherches cytologiques sur les Ascomycétes,” ‘Annales Mycologici, vol. 3, 1905, p. 123. 1905.] Development of the Ascocarp of Humaria granulata. 367 24. Oltmanns, F. ‘“ Ueber die Entwickelung der Perithecien in der Gattung Cheetomium,” ‘ Bot. Zeit.,’ vol. 45, 1887, p. 192. 25. Woronin, M. “Zur Entwicklungsgeschichte des Ascobolus pulcherrimus, Cr. usw.,” ‘ Beitr. zur Morphol. u. Physiol. der Pilze, Zweite Reihe,’ 1866, p. 1. 26. Van Tieghem, Ph. “Culture et développement du Pyronema confluens,” ‘ Bull. de la Soc. Bot. de France,’ vol. 31, 1884, p. 35. EXPLANATION OF PLATES. PLATE 13. Fig. 1.—Portion of mycelial hypha showing the nuclei, and deeply-staining granules on the transverse wall. x 1900. Fics. 24 and 28.—Two young archicarps growing up from the general mycelium : fresh preparations. x 430. Fic. 3.—Slightly older archicarp in section. x 430. Fie. 4.—Archicarp showing the vegetative hyphze beginning to grow out from the cells beneath the ascogonium. The granules on the wall are clearly visible. x 620. Fic. 5.—Young ascocarp in which the ascogonium and the sub-terminal cells of the archicarp are covered in by the vegetative hyphe. x 430. Fig. 6.—Section of young ascogonium showing ascogonial nuclei. x 1050. Fic. 7.—Section of slightly older ascogonium. x 1050. Fic. 8.—Section of ascogonium and basal cell of archicarp which have just become covered in by the vegetative hyphe. The vegetative as well as ascogonial nuclei are clearly visible. x 1050. Puate 14, Fic. 9.—Section of ascocarp of about the age shown in fig. 5. The granules in the ascogonium and the partial emptying of the stalk cells are to be clearly seen. x 430. Fie. 10.—Section of upper part of young ascocarp showing ascogonium partly filled with nuclei and “basal cell” below. Three ascogenous hyphe can be traced throughout their whole length, while portions of others are visible among the vegetative plectenchyma. Nuclei are to be faintly distinguished in the “ basal cell” and some of the vegetative cells. x 1050. Fic. 11.—Section of somewhat older ascocarp showing the ascogonium and three stalk cells. The majority of the nuclei have already migrated from the ascogonium. x 620. Fic. 12.—Section showing the nuclei in the basal cell and the cells immediately surrounding. x 1050. Fic. 13.—Section through the lower part of the ascogonium and the upper part of the basal cell showing the curious granular masses on the transverse wall. x 1010. Fic. 14.—Two female nuclei of the ascogonium in the process of fusion. x 2700. Fic. 15.—A group of ascogonial nuclei with a fusion-nucleus showing nucleoli in act of fusion. x 2700. Fic. 16.—A group of nuclei of ascogonium with a large-fusion nucleus with two nucleoli. x 2700. 368 Sexuality, etc., of the Ascocarp of Humaria granulata. PLATE 15. Fie. 17.—A group of nuclei from an ascogonium showing one fusion-nucleus with two nucleoli, another with nucleoli which are just fusing. x 2700. Fic. 18.—Group of ascogonial nuclei of different sizes. x 2700. Fic. 19.—Ascogenous hyphe and vegetative hyphe showing the distinction of the nuclei. x 1900. Fics. 20—25.—Stages in the development of the ascus at ends of the ascogenous hyphe. Fies. 26—28.—Three stages of the first nuclear division in the ascus. x 1900. Fic, 29.—Spore formation in the ascus. x 1900. Fic. 30.—Section of young ascocarp showing the wedge of paraphyses bursting through the peridium. The nearly empty ascogonium and stalk cells are visible. x 175. Fic. 31.—Section through slightly older ascocarp when the peridium has been burst through completely, the hymenial layer definitely arranged, and a certain number of spores formed. The empty ascogonium is still visible, but the rest of the archicarp has disappeared. x60. [se Roy. Soc. Proc. B. vol. 77 Pl * vy & PTASEr Highley, imp. ith. . Highley, del.et1 eat, ae TLOV. SOG FOC DiVOUL LLL L4 P. Highley, del. et lith. Highley, imp. P Highley, del. et lith. Highley, imp. 369 A Study of the Mechanism of Carbon Assimilation in Green Plants. By Francis L. UsHer and J. H. PriEstLEy, B.Sc., Lecturer in Botany at University College, Bristol. (Communicated by Professor M. W. Travers, F.R.S. Received December 16, 1905,—Read January 18, 1906.) (From the Chemical and Botanical Departments, University College, Bristol.) introduction.—The investigation to be described in this paper has had for its object the elucidation of certain problems concerning the nature of the first stages in the assimilation of carbon from carbon dioxide by the green parts of plants; and although far from complete, it has been thought advisable to publish the results already obtained, inasmuch as the weather is likely to hinder the experimental work for some time to come. In 1870 Baeyer put forward the hypothesis that formic aldehyde is the first product of the decomposition of carbon dioxide in the plant. This suggestion received some support from Bokorny,* who proved in 1891 that starch was formed in the dark by the green filaments of Spirogyra when immersed in a solution of sodium oxymethyl-sulphonate of 0:1 to 1 per cent. strength. Bokorny’s experiments are possibly open to the objection that formaldehyde condenses very readily to non-poisonous carbohydrates in presence of sulphites or bisulphites, and it has been shown by Laurent and Actont that starch is formed in the dark from most sugars. Quite recently, Bouilhac and Trébouxt have succeeded in growing plants _ In a very dilute solution of pure formaldehyde. Tréboux has found that Hlodea forms starch in the dark from a 0:001-per-cent. solution of formal- dehyde, and Bouilhac has shown that this is also the case with Sinapis alba and some Algw. Their experiments bring out in a striking manner the intensely poisonous nature of even very dilute solutions of formaldehyde. Evidence of this kind, however, is quite indirect, and on this account greater importance attaches to the results obtained by Bach,§ who for the first time demonstrated the decomposition of carbon dioxide by light outside the plant. He showed that by passing pure carbon dioxide through a 1:5-per-cent. * ‘Berichte,’ 1891, vol. 24, p. 103. + ‘Roy. Soc. Proc.,’ 1890, vol. 47, p. 150. { ‘Flora,’ 1903, p. 73. § ‘Comptes Rendus,’ 1893, vol. 116, p. 1145. 370 Messrs. F. L. Usher and J. H. Priestley. _[Dec. 16, solution of uranium acetate exposed to sunlight in a glass apparatus, a pre- cipitate consisting of a mixture of uranium peroxide with lower oxides was formed, and that the solution contained formaldehyde. Bach regarded the uranium acetate solution as playing the part of a chemical and an optical sensitiser, and considered the decomposition of the carbon dioxide to result primarily in the production of hydrogen peroxide and formaldehyde. Decomposition of Carbon Dioxide Outside the Plant.—The experiments of Bach have been repeated and confirmed, both as to the production of peroxide and formaldehyde. The amount of decomposition obtained in three weeks in bright weather was extremely small, and this appears to us to be explained by the fact that (1) as a chemical sensitiser uranium acetate is far inferior to that which exists. in a green plant, inasmuch as the separated oxygen (in whatever form it may exist) is not entirely removed from the sphere of action, as in the case of the plant, but remains as a fairly insoluble peroxide which undergoes a rever- sible change with the other product, namely, formaldehyde ; (2) as an optical sensitiser uranium acetate is inferior to chlorophyll to the extent that it . possesses no absorption at all in the red, and only two faint bands between F and G. In view of the extreme slowness of the reaction under these conditions, experiments were made with very large concentrations of carbon dioxide. Tubes of Jena glass, 40 cm. long and 8 to 10 mm. bore, were about three- quarters filled with 1:5 per cent. uranium acetate solution, and cooled in liquid air while some carbon dioxide was passed in. They were then sealed, and suspended outside a south window in bright sunshine. Within 15 minutes of warming up to the air temperature, a precipitate began to form, and in 24 hours the reaction was complete. The tubes when opened were found to contain uranium peroxide and formic acid, but no formaldehyde. The formic acid was obtained by distillation of the filtrate from the peroxide, and was characterised by (1) reduction of silver nitrate, (2) reduction of Fehling’s. solution, and (3) properties of lead salt. Thus with very large concentrations of carbon dioxide, formic acid, and not. formaldehyde, results. These experiments are open to the objection that since uranium acetate 1s: to a considerable extent hydrolysed in solution, the formaldehyde in one case and the formic acid in the other may possibly have been derived from the acetic acid present. It has been found that uranium sulphate in a 2-per-cent. solution functions. in the same way as the acetate. An experiment with the sulphate, conducted in the usual way, 2.¢., bubbling carbon dioxide through the solution, which 1905.| Mechanism of Carbon Assimilation in Green Plants. 371 lasted over three weeks in very dull weather, gave uranium peroxide and formic acid. The different results obtained here from those in the case of similar experiments with the acetate, may be due to the “reduction potential” falling below the limit required for the completion of the second stage of the decomposition. In all the foregoing experiments, except in the case of the liquid carbon dioxide tubes, blank experiments were simultaneously performed, (1) with uranium solution and carbon dioxide in the dark, and (2) with carbon dioxide free solution in the light. In neither case was any precipitate formed. Decomposition of Carbon Diowide in the Plant—I{ a similar reaction, resulting in the formation of formaldehyde and a peroxide, takes place in the first stage of the absorption of carbon dioxide by the plant, it is obvious that both the initial products of decomposition must undergo a rapid change. On account of its intensely poisonous nature, formaldehyde must be very rapidly converted into some physiologically inert substance ; and the peroxide must be decomposed with evolution of gaseous oxygen, a process which follows exposure to light by an interval of one or two seconds. The problem, then, is to ascertain the process by which oxygen is dis- engaged; to show the actual presence of formaldehyde localised in the neighbourhood of the chloroplasts; and to trace the steps by which the formaldehyde is polymerised. The Mechanism of the Evolution of Oxygen from the Green Plant.—In the experiments relating to the decomposition of carbon dioxide outside the plant, no evolution of oxygen gas is ever observed; it remains in the system as a peroxide. There have been conflicting statements with regard to the presence of hydrogen peroxide in plants, but even if traces are to be found, there is no evidence that it 1s a product of decomposition of carbon dioxide. It has indeed been shown that several organic substances, notably the organic acids, é.g., oxalic, give rise to hydrogen peroxide on exposure to light, and such substances as these are of common occurrence in the leaves of plants. If, however, hydrogen peroxide is one of the first products of the photolysis of carbon dioxide, we are more directly concerned with the elimination of oxygen in the gaseous form than with the detection of the peroxide. Hitherto those writers who have recognised the difficulty at all have suggested some method of reduction, which, of course, leads back to the starting point. It appeared much more probable that this step in the process was brought about by a catalyst, probably an enzyme. To test this, some Elodea was immersed in a dilute solution of hydrogen peroxide. An 372 Messrs, F. L. Usher and J. H. Priestley. _[Dee. 16, immediate and rapid decomposition set in, and a gas was freely evolved, which was found to be oxygen. The action proceeded as rapidly in the dark as in the light. The following experiments were performed with the object of ascertaining the nature of the catalyst :— (a) A plant was immersed in boiling water for 50 seconds and was sub- sequently found to be without action on hydrogen peroxide. (b) After treatment with dilute solutions of iodine, mercuric chloride, and hydrogen sulphide, no action took place. (c) Some Elodea was suspended in air charged with chloroform vapour for two hours in order to kill the protoplasm, and was then allowed to “ recover ” for a similar period. Rapid disengagement of oxygen took place. (d) After immersion in very dilute formaldehyde solution, hydrogen peroxide was not decomposed. These experiments seem to point to the existence of a catalysing enzyme. Several attempts to extract it by simple maceration with water or salt solution failed, and we were also unable to extract it after powdering leaves in liquid air. Following a suggestion of Dr. Horace Brown, we ultimately succeeded in obtaining it by previously drying a quantity of Hlodea, and subsequently digesting with water at 30° for 48 hours. The enzyme was precipitated by an excess of absolute alcohol and dried. By this process it is obtained as,a light brown powder, containing diastase, whose aqueous solution energetically decomposes hydrogen peroxide, whereas ordinary malt diastase does not. Whether the enzyme is one already known, or whether it is secreted specially for the purpose of catalysing. hydrogen peroxide, we cannot as yet say. On mounting a leaf of Hlodea in very dilute hydrogen peroxide, and examining microscopically under a high power, bubbles of gas were seen to emerge from the chloroplasts only, an observation which shows the strict localization of this enzyme to the seat of the photosynthetic process. In regard to the distribution of this enzyme, we have examined the foliage leaves of plants belonging to 46 Natural Orders and representative of the Vascular Cryptogams and all the main groups of the Phanerogams, and have found the power of catalysing hydrogen peroxide in every case, though the energy of the decomposition varies considerably in different groups. It also occurs in etiolated leaves and in potato tubers, and, in fact, appears to be associated with amyloplasts, whether possessing chlorophyll or not. The Production of Formaldehyde and the Manner of its Removal_—tit has been found in the case of Spirogyra that starch appears in a previously 1905.| Mechanism of Carbon Assimilation in Green Plants, 373 starchless filament within three minutes of exposure to light, and it is probable that some sort of carbohydrate is formed much sooner than this, for it has been shown by Brown and Morris* that starch is probably§not elaborated within the cell until the supply of nutriment is in excess of the cell requirements. It would therefore seem as though the arrangement which exists in the plant for the removal of formaldehyde is at least as efficient as any external arrangement we can make to remove it in a different way, without at the same time killing the plant, and thus eliminating one of the essential factors, namely, the vitality of the protoplasm. For this reason it is useless to look for formaldehyde in healthy assimilating leaves. It is well known that certain chemical substances possess the property of condensing formaldehyde to various carbohydrates, chiefly formose, a-acrose, and methylenitan. It has been found by Loewt that such condensing agents are chiefly metallic oxides and acid sulphites, substances not likely to occur in plants. Moreover, condensation by these bodies is a comparatively slow process, and quite inefficient when applied to the requirements of a plant. Nevertheless, if the condensation in the plant were due to some chemical agent stored in the neighbourhood of the chloroplast, it should still be capable of taking place when the protoplasm of the leaf is killed and its enzymes destroyed. Some healthy green sprigs of Hlodea were immersed in boiling water for 30 seconds, in order to kill the protoplasm and destroy the enzymes. They were then placed in water saturated with carbon dioxide and exposed to sunlight. In the course of a few hours the deep green colour of the leaves had been completely bleached, and on immersing the bleached sprigs in a solution of rosaniline decolourised with sulphurous acid, a red colour was developed. The original green material when treated in this way exhibited no colouration. There was, therefore, some substance of an aldehydic nature present in the killed and bleached leaves which was absent in those which were alive. The sequence of events in this experiment may be described as follows :—Photolysis of carbon dioxide begins in the normal way, giving rise to hydrogen peroxide and formaldehyde. The enzymes having been destroyed, the hydrogen peroxide, instead of being catalysed in the usual manner, oxidises the chlorophyll to a colourless substance, at which point the reaction necessarily comes to an end. Meanwhile a quantity of * ©J.C.S.,’ 1893, ‘ Trans.,’ p. 632. + ‘Berichte,’ 1888, p. 271. 374 Messrs. F. L. Usher and J. H. Priestley. [Dee. 16, formaldehyde, equivalent to the hydrogen peroxide required to destroy the chlorophyll, accumulates, and thenceforward the reaction is strictly reversible. / The following experiments were performed to settle the points involved in this explanation :—It was in the first place necessary to show whether the colouration referred to above was due to formaldehyde. For this purpose some leaves, killed and bleached in carbon dioxide solution as described, were soaked for 12 hours in aniline water, and were then examined micro- séopically under a high power. Some leaves which had been killed and simply decolourised with hydrogen peroxide were treated in the same way. In the first case the decolourised chloroplasts were observed to be the centres of clusters of well-defined crystals, identical in appearance with those of methylene aniline, artificially prepared from aniline water and form- aldehyde. They were soluble in dilute mineral acids and also in warm alcohol, from which they crystallised in the cell on cooling. The leaves artificially decolourised with hydrogen peroxide showed no crystals. An attempt was then made to obtain the formaldehyde outside the plant. For this purpose a large quantity of Ulva and Enteromorpha was killed and bleached in carbon dioxide solution, and subjected to steam distillation. The distillate was divided into two parts. To the larger of these was added some aniline water. A white precipitate was formed after some time, which was collected, and heated side by side with a comparison tube containing methylene aniline. It melted, not quite sharply, three or four degrees below the pure artificially prepared substance. The other portion of the distillate was evaporated with -ammonia on the water-bath, and the residue dissolved in water and treated with bromine water, gave the characteristic tetra-brom derivative of hexamethylene-tetramine. Hence, leaves in which both protoplasm and enzymes have been killed, when placed under conditions favourable for assimilation, develop formaldehyde, until the photolytic process is brought to an end by the destruction of the chlorophyll. It was next necessary to determine whether the condensation of the formaldehyde is due to an enzyme secreted by the chloroplast, or whether the protoplasm of the granule itself effected it. Some H/odea was suspended in air charged with chloroform vapour for two hours, by which means the protoplasm was killed without affecting the enzymes. It was then exposed to sunlight in saturated carbon dioxide solution. In a few hours the chlorophyll became bleached, and formaldehyde was subsequently found in the plant. 1905.| Mechanism of Carbon Assimilation in Green Plants. 375 It follows from this that the protoplasm of the chloroplast is the con- densing agent. The bleaching of the chlorophyll in this case is due to the fact that the enzyme, though unharmed at the commencement of the experiment, quickly becomes poisoned by the accumulating formaldehyde. In this last experiment the limit to the accumulation of formaldehyde in a plant is realised, since there is formed an amount equivalent to that amount: of hydrogen peroxide which is catalysed before the enzyme ceases to act together with an amount equivalent to the hydrogen peroxide required to destroy the chlorophyll. That a certain amount of hydrogen peroxide is catalysed when the protoplasm only is dead was shown in the following manner. Approxi- mately equal quantities of Hlodea were taken, one of which (A) was killed by immersion in boiling water, another (B) was suspended for two hours in air saturated with chloroform vapour to kill the protoplasm and not the enzymes, while the third (C) served as a control. All three were placed in carbon dioxide solution under funnels with inverted test-tubes, and exposed to artificial light for 12 hours. From A there was no evolution of oxygen, from B 0°3 e.c. were given off, and from C 28 c.c. In the case of B, after exposure to light, no catalytic action on hydrogen peroxide could be observed, while previously vigorous decomposition had taken place. Summary. 1. The photolysis of carbon dioxide may take place outside the plant in absence of chlorophyll, provided one of the products is removed. 2. The normal products of the photolysis are hydrogen peroxide and formaldehyde, though under certain conditions formic acid may be formed. 3. In the plant the decomposition of the hydrogen peroxide is provided for by a catalysing enzyme of general occurrence. 4. The condensation of the formaldehyde is dependent on the healthy condition of the protoplasm. There are therefore three factors essential to photosynthesis from carbon dioxide and water in the plant, they are (i) vitality of the protoplasm, (11) presence of a catalysing enzyme, and (iii) presence of chlorophyll. If any one of these factors be interfered with, the process of photosynthesis ultimately comes to an end, through the destruction of the optical seusitiser, chlorophyll. LXXVII.—B, 25 376 Mechanism of Carbon Assimilation in Green Plants. The relations between the various factors in this process may be diagram- matically expressed thus :—— Carbon dioxide + Water we —_ = | [f not removed, destroys|= CHLOROPHYLL h if = Hydrogen peroxide + Formaldehyde [If not removed, poisons | x ENZYME LivIn@ PROTOPLASM Oxygen Carbohydrates In conclusion, we wish to express our indebtedness to Dr. Travers, and to Dr. Horace Brown, for their valuable suggestions and help in the course of this research. 377 A Biometrical Study of Conjugation in Paramecium. By Raymonp Pearz, Ph.D. (Communicated by Professor Karl Pearson, F.R.S. Received November 15,— Read December 7, 1905.) (Abstract.) 1. A study of variation and correlation in conjugating and non-conjugating specimens of the common ciliate infusorian, Paramecium caudatum, was undertaken for the purpose of obtaining answers to the following questions :— a. Is the portion of the Paramecium population which is in a state of conjugation at a given time differentiated in respect of type or variability or both from the non-conjugating portion of the population living in the same culture at the same time ? b. Is there any tendency for like to pair with like (“ homogamy ”) in the conjugation of Parameecium, and if so, how strong is this tendency ? The material on which this paper is based is comprised in eight series, taken from three different cultures at different times, and includes altogether the measurements of 1894 individual Paramcia. The characters studied were length and greatest breadth of the body, length-breadth index, and the difference in Jength between the two individuals of a pair of conjugants. In the measuring conjugant pairs were taken quite at random, and then in each case the two undistorted non-conjugant individuals which were lying nearest in the field of view of the microscope to the conjugant pair were measured. This procedure was followed to avoid any sub-conscious bias in choosing non-conjugants. The cultural history of the different series may be summarised as follows :— Series A, C, D, and #.—These series all came from a single culture in the Zoologisches Institut, at Leipzig. This culture was set with dry hay and pond water, July 25,1905. The dates of collection and measurement and number of individuals in each series are given in the table on p. 378. Series B.—This series came from another culture at Leipzig set in the same manner as the one just mentioned. Conjugants were found on August 22, but in very small numbers. On the next day only two pairs of conjugants were found, and after that none at all. So that, all told, Series B included only 12 pairs of conjugants and 24 non-conjugants. Series AA, Fy, and Fy.—These series included only conjugants. ‘They were 378 Dr. R. Pearl. A Biometrical [Nov. 15, measured from material in the Zoological Laboratory of the University of Michigan, collected by Professor D. C. Worcester. They all came from a single culture set with decaying plant material and pond water. Series AA includes 200 pairs of conjugants, Series Fy 70 pairs, and Series Fy, 77 pairs. Number Number of Series. Dates of measurement. of conjugants | non-conjugants measured. measured. JN se5008 August 15, p.m.—August 18, noon, inclusive ...... 105 pairs 210 Cae ce: » 24, AM.— ,, 26, P.M., se asaBeo ioxl Fs, 202 DE enn Fyatt BOLI dae seescbaceters soutcotseretnesetedmadectvasioas UGS 32 I Boc0e aSeptemiberiGicsce-0,-sercceaceceteaeaaysties-eesees eer eee eee _ 132 For further details regarding the measurements, culture histories, etc., the complete paper must be consulted. 2. An examination of the variation constants shows that Paramecium is relatively slightly more variable in breadth than in length of body, though the difference is not large. For the variation in length the coefficients of variability for different series (including several other long series besides those collected in this work) are found to cluster well together about a value of 8 to 9 per cent. This is a much lower value than has been found by other workers{ for variation in similar size characters in organisms with firm exoskeletons. In the characters studied Paramecium follows the same general laws which have been shown to hold for continuous variation in higher forms. 3. It was found that conjugants are markedly differentiated from non- conjugants living in the same culture in both type and variability. This differentiation includes all the characters studied. An idea of its extent and direction may be gained by an examination of the following table. In it are given (a) the absolute differences between conjugants and non- conjugants in respect to the character and constant designated; (?) the probable errors of these absolute differences; (¢) the relative differences defined as the percentage which the absolute difference is of the non- conjugant constant in each case. The absolute differences are taken as positive when the non-conjugant constant is greater. Only two series (A and C) are taken here as illustrations. In the complete paper similar data for other series are given in detail. * Series D and B (vide infra) include only a few individuals, because at the time they were collected no more conjugants were to be found in the cultures from which they came. + For detailed references, see complete paper. 1905. | Study of Conjugation in Paramecium. 379 Table I.—Differentiation of Conjugants from Non-conjugants. | Ore Relative differ- ] Absolute difference eGR Series. | Character. Constant. berween Pon sboningants non-conjugants JOBE. and conjugants. Per cent. A Length ...... NIGER. ocecboedoadanshoesben 21 °833 +0°893 micron 11°5 my) Pe REE er Standard deviation 4,337 +0°631 53 27:°9 0 mp. toBd400 Coefficient of variation 1°517 +0349 per cent. 18 °5 a Breadth...... Means vasieas oisene scene 8°456 +0°335 micron 16:0 5 TRE aes Standard deviation ... 1°700 +0°237 3 28 -96 5 Toni tesa Coefficient of variation} 1°714 +0484 per cent. 15 ‘4 , Iniclexveasessse Mieaniens tacecucsecsaactoes 1510 +0°171 2 54 y) OA Uae aeaiiel Standard deviation ...| —0°195*+0°121 a ate s Length and} Coefficient of correla- 0 °3107 +0 °0526 52°7 breadth tion C Length ...... Meam \scaicceacieecoonaorss 33 341 +1°098 micron 15°9 Ss Regence Standard deviation ...) 6°005 +0°777 5 31 “4 Fh Fi sede Coefficient of variation 1°684 +0°398 per cent. 18 °5 rf Breadth...... MC aM s sriseciesnesanessteaes 11 °050 +0324 micron 20 ‘4 ) opi) cconaon Standard deviation ...) 2°491 +0°229 55 42 *2 90) Ot ecadag Coefficient of variation 2°984 +0°456 per cent. 27 “A p TeaVelOs |," 86:0 Not acid : | determined | Table II. | de \\valyly? Zeist 8 | 4. ee es ee S PALCOHOM ae ececsccce tae seanencssemenien 1°34 1-28 1-43 1:01 | AGE Sa5osanconssqesooas0de300~ 0°31 0°25 0 52 1°13 | Lacticacid .-.-....... 0... cose one 0°33 0°28 0°55 91! |eStccinic acid’ 20)... ..c-eeneceseseees 0-15 0°19 0-27 0°32 | Formic acid ...........2...:.0.seeceeee 0 04 0-02 0:07 0-00 Carbon) dioxides.s---.2--00--ees-ee-ess 1°60 — 1°44 0°74 otal urs. ee 3°77 | o- | 4°28 | 511 Hydrogen, atoms per molecule ...| 1°33 | — | 1-50 | Wey It will be observed that the ratio of hydrogen to carbon dioxide by volume is about 0°5 to 1, whilst these gases are produced by JB. colt communis in approximately equal volumes. Theobald Smith,* using an ordinary fermentation tube, gives the characteristic ratio for B. coli communis as H2/CO2 = 2:1 and for B. lactis aerogenes H2/CO2=1:1. This difference is due to the solubility of the carbon dioxide in the liquid medium, and it must be remembered that while Smith’s ratios give a perfectly satisfactory working test for the discrimination of the organisms, they do not represent the actual volumes or ratios of the gases produced. Further examination of the fermentation products revealed the fact that * Loe. cit. 1905.] __ B. lactis aerogenes on Glucose and Mannitol. 401 no other acids had been formed, and search was therefore made for com- pounds of a different type. It was previously suggested that the deficiency of carbon observed in the fermentation produced by L. coli communis, amounting to only 0°25 to 0-9 of an atomic proportion of carbon, might possibly be due to the presence of reduction products of sugar, and com- pounds of this kind were therefore sought. It was found that when the neutral liquid, containing the products of fermentation along with peptone, was evaporated to dryness at 55° under diminished pressure and extracted with alcohol, a solution was obtained which yielded on fractionation a colourless liquid boiling at 181° to 183° (corr.) at 760 mm. pressure. The yield was very small, only amounting to about 1 gramme per litre of medium containing 20 grammes of glucose, but it was found possible to increase the yield by employing a medium containing 5 per cent. of glucose, and in this way 8 grammes of the new substance, containing 52°8 per cent. of carbon, were obtained per litre of medium contaiming 50 grammes of sugar. This only accounts for about two-thirds of the missing carbon, and a rough estimate of the amount lost during the process of distillation and extraction was, therefore, made by dissolving 8 grammes of the material in 500 c.c. of a medium containing 5 grammes of Witte peptone, 6 grammes of calcium lactate and 65 grammes of alcohol and then extracting it in the manner described above. Only a2 grammes were recovered, the loss per 500 ec. being therefore about 2°$ grammes and the loss per litre about 56 grammes. This brings the total amount produced from 50 grammes of glucose to about 13°6 grammes, slightly in excess of that required. It is hoped that the actual yield may be increased by a careful fractionation of the fermentation products. The new product is apparently a mixture, and it has not yet been found possible to separate and identify all the components, so that the following must be taken as only a preliminary account of the substance. It boils at 181° to 183° (corr.), and solidifies in the cold to a transparent mass which melts indefinitely at about 28°. It is optically active, the value for [¢]p for different preparations varying from 0-46 to 0-71. The com- position of the substance dried by quicklime is approximately that of a butyleneglycol, but the percentage of carbon is about 0°6 too low. It does not reduce Fehling’s solution either in the cold or on heating. That this substance contains a large proportion of 2:3-butyleneglycol, CH;.CH(OH).CH(OH).CHs, is shown by the following facts :— 1. When the liquid is heated with phenylisocyanate dissolved in anhydrous ether, combination occurs and a mixture of urethanes is produced. The fraction of these which is least soluble in aleohol comprises about 90 per cent. 262 402 Dr. Harden and Mr. Walpole. Chemical Action of [Dee. 5, of the whole amount and has the composition of the diphenylurethane of butyleneglycol (CsHy)02.2CsH;NCO) :-— Analysis. Found. Calculated. CEO Res: | 65°79 65°85 Le anatase ae 6°21 6°09 Nets ont 8:57 8°35 It is sparingly soluble in cold alcohol, ether and benzene, crystallises in rosettes of needles and melts at 197° to 198° (uncorr.). When the urethane is boiled with baryta water or caustic soda solution it is decomposed and yields a glycol boiling at about the same temperature as the original material. This glycol has, however, not yet been isolated in the perfectly pure and dry state. A monourethane, CsH,)02CsH;NCO, has also been prepared which is somewhat more soluble in cold alcohol than the diurethane, and crystallises. in needles, melting at 100°:— Analysis. Found. Calculated. IN GS ae eeeeene 6°89 6°65 2. Both the crude glycol and that recovered from the diurethane are con- verted by oxidation with bromine in the light* into diacetyl, CH;.CO.CO.CH;. which was recognised by its extremely characteristic appearance and smell, and by the formation of a phenylosazone melting at 242° to 242°8 (uncorr.).+ The formation of this substance shows conclusively that 2 : 3-butyleneglycol must be present in the fermentation product. Detection of Acetyliethylearbinel among the Ferinentation Products—1t was further found that the distillate from the liquid in which the organism was grown reduced Fehling’s solution in the cold and gave with phenylhydrazine the osazone of diacetyl, melting at 243°. These properties point to the presence in the distillate of acetylmethylearbinol, CH3.CO.CH(OH).CHs, which has previously been detected in this way by Grimbert? and by Desmots§ in the products of the fermentation of glucose by several bacteria = B. tartricus, B. mesentericus vulgatus, B. fuscus, B. flavus, B. niger, B. ruber, B. subtilis, and Tyrothrix tenuis. It has also been found in vinegar. This compound appears only to be formed in very small amount. Since it * y. Pechmann, ‘ Ber.,’ 1890, vol. 23, p. 2427. v. Pechmann, ‘ Ber.,’ 1888, vol. 21, p. 2754. ‘Compt. Rend.,’ 1901, vol. 132, p. 706. ‘Compt. Rend.,’ 1904, vol. 138, p. 581. GK ++ 1905. | B. lactis aerogenes on Glucose and Mannitol. 403 is likewise converted into diacetyl by oxidation with bromine in the light, it is important to notice that the glycol used for conversion into diacetyl], as described above, was quite free from any substance capable of reducing Fehling’s solution, and yielded a relatively large amount of diacetyl. Il. Action of B. lactis aerogenes on Mannitol. A quantitative examination of the products of fermentation of mannitol by B. lactis acregenes showed that in this case also the action differed from that produced by B. coli communis, but that the deficit of carbon was only one-half of that found for glucose. This is shown in the following tables: Table III giving the percentages and Table IV the number of carbon atoms per molecule of mannitol represented by the products in two experiments (Cols. 1 and 2). As before, the products obtained by the action of B. coli communis are also given for the sake of comparison (Col. 3) :— Table IIT. ] 2: 3 PAS COMO Ree: ei ataciriess tse crac nacaeeweess 32°5 32°35 281 eAtceticraciduuns. passe dacissuacnensennss 2°5 2:1 9°5 Malcticlacidhee gt: .: cccosassteeauteens 8°6 8°6 18 6 DEANE EXO! LoonssoncaosasoononEobed 3:2 2°8 8:9 ING ETC EYENGl ao onosossboSocsabeobdac iS 1°6 3°0 Carbon dioxide..........0cecsseeeeeee 3DEo) 35 °5 28 44 Carbon dioxide, c.c. per gramme} 180°3 180 °3 143 -0 Hydrogen, ¢.c. per gramme ...... 138 °3 143 6 167 0 TOs Gy OXO), Soccocooopoponboodocgnen 0°77 0:79 1:18 Percentage excess of 1-lactic acid 65-0 56 0 79-0 J Table IV. | 1. | 2 | 3. ACOH OMA Rie es metaren tect erciens 2°57 Jeo" 2°22 BATCETIC ACL sa vase Suis cae ce canaeneete 0°15 0:12 0°58 Macticvacid! eta. swccnsacaceekeeceecees 0°52 0°52 1°13 SHC OUAG GO!” cooononsoncnnooeooenboCo 0-20 0°17 0°55 HOnmic acids ssaecee see 0:06 0 :065 0:12 Carloon dioxide...............0.c0 eee 147 1°47 1°16 Totally saiscecc | 4:97 491 5°76 | H atoms per molecule glucose al 2-26 | 2°34 | 2°7 | PVs why | 404 Action of B. lactis aerogenes on Glucose and Mannitol. . Further examination has shown that in this case as in that of glucose, both acetylmethylearbinol and a glycol are produced, but both in much less quantity. The amount of crude glycol actually isolated from the products of fermentation of 50 grammes of mannitol was only 0°75 gramme. Since, however, the loss in isolating may be roughly taken as about 5 grammes, this is approximately the yield which would be expected if 6 to 7 grammes were formed. The nature of these products and their quantitative estimation, as well as the study of their optical properties, is still under investigation, and search is also being made for these and similar substances among the fermen- tation products of other bacteria. General Considcrations. The production of so large a proportion of 2:5-butyleneglycol in these experiments affords clear proof that this substance is derived from the glucose. The interesting question as to the mode of its production from the glucose or mannitol molecule will be best deferred until a more complete examination of the products, and especially of their optical relations, has been made. The close constitutional relation between the glycol and lactic acid, and the readiness with which its oxidation product —diacety1—passes into an aromatic compound are also points of great interest. It may, however, be noted that the comparison of the fermentation products of Bb. coli communis and B. lactis aerogenes shows, firstly, that the alcohol produced by the latter organism is slightly greater in amount than that due to the former, and, secondly, that it is at the expense of that part of the molecule which in the B. coli fermentation yields acetic.acid and lactic acid, that the B. lactis aerogenes forms the new products. It may further be observed that both these bacteria produce twice as much alcohol from mannitol as from glucose, a fact which tends to confirm the suggestion previously made,* that the formation of alcohol in these reactions is related to the presence of the terminal CH.(OH).CH(OH) group, which occurs twice in the molecule of mannitol and only once in that of glucose. A substance of the composition of butyleneglycol has previously been isolated from the products of fermentation of sugar by yeast,f and was also found in winef and. in brandy.§ This substance, boiled at 178° to 179°, yielded a diacetin boiling at 192° to 193°, and was considered to be identical * Harden, ‘ Trans. Chem. Soc.,’ 1901, p. 601. + Claudon and Morin, ‘ Compt. Rend.,’ 1887, vol. 104, p. 1109 ; Henninger and Sanson, ‘Compt. Rend.,’ 1888, vol. 106, p. 208. { Henninger, ‘Compt. Rend.,’ 1882, vol. 95, p. 94. § Morin, ‘Compt. Rend.,’ 1887, vol. 105, p. 1019. The Alcoholic Ferment of Yeast-Juice. 405 with the synthetical isobutyleneglycol of Nevolé,* which boils at 176° to 178°. The yield obtained from sugar was, however, very small, and only amounted to about 0:2 per cent. after allowing for the losses involved in the extraction of the compound, In view of the properties of the crude glycol described above, it would seem advisable to re-examine Henningevr’s glycol, the constitution of which was not experimentally examined. The Alcoholic Ferment of Yeast-Juice. By Artuur HarpeEn, D.Sc., Ph.D., and WILLIAM JoHn Youne, M.Sc. (Communicated es Dr. C. J. Martin, F.R.S. Received December 8, 1905,—Read February 1, 1906.) (From the Chemical Laboratory, Lister Institute.) 1. Effect of the addition of Boiled and Filtered Yeast-jwice on the Fermentation of Glucose Produced by Yeast-juice. In the course of some experiments on the action of various proteids on the fermentative activity of yeast-juice, it was observed that the alcoholic fermentation of glucose by yeast-juice is greatly increased by the addition of yeast-juice which has been boiled and filtered, either when fresh or after having undergone autolysis, although this boiled liquid is itself incapable of setting up fermentation. Thus, the total fermentation produced by yeast- juice acting on excess of glucose is, as a rule, doubled by the addition of an equal volume of the boiled juice, and a further increase is produced when a greater volume is added, the sugar concentration being kept constant.f A similar observation was previously made by Buchner and Rappj in a single experiment (No, 265). The following table embodies a few of the results obtained, the yeast-juice being prepared and the amount of carbon dioxide evolved being estimated by * “Compt. Rend.,’ 1876, vol. 83, p. 65. + Harden and Young, Preliminary Note, ‘Proc. Physiol. Soc.’ 1904, vol. 32, November 12. { ‘ Ber.,’ 1899, vol. 32, p. 2093. 406 Dr. A. Harden and Mr. W. J. Young. [ Dec. 8, the method previously employed by the authors.* In every case the con- centration of sugar was kept constant, and both in these and all the fermentation experiments described in this paper, toluene was added as an antiseptic. ce Table I.—Effect of the Addition of Boiled Yeast-juice on the Total Fermentation of Glucose by Yeast-juice. | No. | Juice. Water. ae Glucose. Time. eae C.e. C.c. c.c, grammes. hours. gramme. 1 25 25 (0) 5 72 0-137 | a oe 25 (0) 25 5 72 0 °378 2 20 20 (0) 4 44 0°115 pas 20 (0) 20 4 44, 0 °363 3 25 0) (0) 2°5 40 0 °370 Eo ae 25 (0) 25 5 40 0-620 d 20 40 0 6 42 0-458 | aad 20 0) 40 6 42 0 °858 " 25 25 0 5 44, 0 °346 AE 1 25 0) 25 5 44, 0-709 6 25 25 0 5 48 0°110 ee 25 0 25 5 48 0-216 | 7 25 25 0 5 60 0°273 | Bees 25 (0) 25 5 60 O -466 | 8 f{ 25 25 0) 5 120 0-424 spate | 25 0 25 5 120 0-959 25 25 @) 5 72 0 °414 25 20 5 5 72 0-546 Queers < 25 10 15 5 72 0°735 | 25 5 20 5 72 0°810 25 0 25 5 72 0-924 (25 25 ny) 5 70 0-246 25 O 25 5 70 0 °356 10 3 25 50 0 VES 70 0-180 sores | 25 0 50 7°5 70 0-431 | 25 75 ) 10 70 0-141 | 25 (0) 15 10 70 0°515 In Experiments 1 to 5 the juice added had been autolysed before being boiled; in Nos. 6 to 8 the added juice was boiled as soon as it had been prepared. Experiments 9 and 10 show that each successive addition of boiled juice, from 0-2 to 3 volumes, produces a further increase in the amount of the fermentation. A similar effect is produced, (1) By the precipitate produced in boiled yeast- juice by the addition of 3 volumes of alcohol (Experiment 1, Table II); (2) By the liquid formed by the autoplasmolysis of yeast, when it is allowed to stand at the air temp2rature for some time (Experiments 2 and 3, Table IT) ; (3) By the liquid obtained by boiling Buchner’s “ Aceton-Dauerhefe” with * Harden and Young, ‘ Ber.,’ 1904, vol. 37, p. 1052. 1905. | The Alcoholic Ferment of Yeast-Juace. 407 water (Experiment 4, Table II). Further, yeast killed by acetone and ether (Aceton-Dauerhefe) reacts with boiled juice in the same way as does yeast- juice (Experiment 5, Table IT). Table II.—Effeet of Various Substances in Increasing Alcoholic Fermentation. i a Uae Ween : lee Sa! ss _ | Yeast- Ke ; ; Carbon No, ice! Addition. Glucose. Time. | dioxide. | | 7 ae Ol ; | ; Ci: ; hours. gramme. 1 25 713) CYGn WERE pecobocoooccoccunbocccdEe0 3) 48 0-110 25 | 25 c¢.c. water + precipitate by 5 48 0-268 _ 75 per cent. alcohol from 25 c.c. . boiled fresh juice 25 Filtrate from 25 e.c. boiled fresh 5 48 O-'141 juice + 3 volumes alcohol, made | to 25 c.c. 25 25 c.c. water + precipitate from 5 48 0-286 25 e.c. boiled old juice by 75 per cent. alcohol | i | 2 25 OS GGL IEE pant ocosoban coh ane ebabeccns 5 | 72 0-070 25 25 c.c. autoplasmolysed yeast- 5 72 0-189 juice, made neutral Ee eae TDG I0.c. water es cco les. tees vessces 5 72 0-084 25 25 ¢.c. autoplasmolysed yeast- 5) 72 0-172 juice, made neutral DOW A 2otGIG! waters. f.sccankccctas sed: cecedos 5 72 0°475 4 25 | 25 ¢.c. aqueous infusion of 5 72 | 0-625 2 grammes Aceton-Dauerhefe | 5 |2 grammes) 40 c.c. water .........:.ccceeecede eee ens 4 48 | 0:062 Aceton- | Dauerhefe | " | 20 c.c. water + 20 c.c. boiled juice | 4. | 48 0-136 2. Dialysis of the Boiled Juice. The constituent of the boiled and filtered juice to which this effect is due is removed when the liquid is’ dialysed in a parchment tube, leaving an inactive residue. In the experiments detailed in the following table (Table III, Experiments 1, 2 and 3) the effect of the addition of boiled juice is compared with that produced by the residue and dialysate respectively. In Experiment 4, the unboiled juice was dialysed, and the fact that the dialysate had a similar ettect to a boiled juice shows that the active constituent exists in the original yeast-juice and is not formed during the boiling. 408 Dr. A. Harden and Mr. W. J. Young. [Dec. 8, Table III—Dialysis of Boiled Yeast-juice. 25 cc. yeast-juice+5 grammes glucose + toluene. | No. | Water. ola Residue. Dialysate. | Dime, Carbon | juice dioxide. c.c. C.c. c.c c.c. hours. gramme. 1 25 ) ) i) 48 0-253 @) 25 (0) 0 48 0°561 0 0 25 oO 48 0 :264 2 25 0 (6) 0 4s 0-268 0 25 i) ) 48 0-497 i) ) 25 0 48 0-276 Verse en 25 ft) 0 0 72 0-113 ) 25 0 0 72 0 +334. i) 0 25 ) 72 0-189 c 0 0 25 72 0-334 | 4 25 0 0 0 48 0-154 i) ) i) 25 48 0-251 3. Dialysis of Yeast-juice. The facts above detailed suggested the possibility of dividing yeast-juice into two fractions by dialysis; an inactive residue and a dialysate which, although itself inert, would be capable of rendering this residue active. This was experimentally realised by filtering the juice through a Martin gelatin filter.* This method of rapid dialysis was chosen because the yeast-juices at our disposal lost their activity too rapidly to permit of the ordinary process of dialysis through parchment being carried out. Hither a 10- or a 7:5-per-cent. solution of gelatin was used to impregnate the Chamberland filter and the filtration was carried out under a pressure of 50 atmospheres. Only a portion of the juice placed in the filter was actually filtered, the remainder being simply poured out of the case as soon as a sufficient quantity of filtrate had passed through. The residue adhering to the candle, which consisted of a brown viscid mass, was dissolved in water and made up to the volume of the juice filtered. Glucose was then added and one portion incubated at 25° with an equal volume of sugar solution and a second portion with an equal volume of the filtrate or of a boiled juice, containing an equal amount of glucose. Before incubation the carbon dioxide was pumped out of all the solutions. The filtrate was invariably found to be quite devoid of fermenting power, none of the enzyme having passed through the gelatin. * “Journ. Physiol.,’ 1896, vol. 20, p. 364. 1905. | The Alcoholic Ferment of Yeast-Jwice, 409 The results (fable [V) show that in this way an almost inactive residue can be obtained which is rendered active by the addition of the filtrate (Experi- ments 1, 2, 3) or a boiled juice (Experiment 4). Table [V.—Filtration of Yeast-juice through the Martin Gelatin Filter. 1d.¢.c. residue+3 grammes glucose + toluene. No. | Water. | Filtrate. | Boiled juice, Time. Carbon dioxide. | c.c. ° cc. c.c. hours. | gramme 1 15 0 0 i G45. 0-000 0 15 0 Fee | 0-035 2 15 0 0 60 | 0-001 0 15 0 liner TEOMA | 0-051 3 15 0 0 60r| 0-008 0 15 0 60 | 0-064 4 15 0 0 60 | 0-024 0 0 15 60 | 0-282 The total fermentations observed even in the presence of the filtrate are very low, this being, at all events in part, due to the fact that in this series of experiments the original juices themselves happened to be of low fermenting power. In a second set of experiments (Table V) a smaller quantity of juice was placed in the filter and the filtration was continued until no more liquid would pass through. The residue was then washed several times by adding water and forcing it through the filter. The time occupied in this process varied greatly with different juices, the limits for the filtration and washing of 50 cc. of juice, using two filters simultaneously, were about 6 to 12 hours. The carbon dioxide was not estimated by absorption in potash as in the previous cases, but was collected and measured over mercury, by means of the apparatus described later on, the object of this procedure being to ascertain not only the total amount of carbon dioxide produced, but the rate and duration of the evolution. The residue was dissolved in water and made to the same volume as the originai juice, and the filtrate was evaporated down to the same volume. All the solutions were saturated with carbon dioxide at the temperature of the bath (25°) before the measurements were commenced, and the observations were continued until all fermentation had ceased. The boiled juice added in Experiments 1, 3 and 4 (Table V) was obtained. by boiling a portion of the same preparation as was used for the filtration. The carbon dioxide is expressed in cubic centimetres under atmospheric conditions. 410 Dr. A. Harden and Mr. W. J. Young. [ Dec. 8, Vol. of i ue yen Boiled i Carbon No. juice filtered. Wash water.| Residue. | Filtrate. juice. Glucose. dicde: c.c. c.c. c.c .c. c.c. | grammes, C.c. 1 75 200 25 0 (0) | 25 10 °4 20 (0) 25 5 396 °3 2 80 ‘260 20 (0) (0) 2 8:3 20 20 (0) A 90 -2 e 3 100 250 25 (0) (0) 2°5 0°4 25 (0) 25 5 268 A 50 200 25 @) (0) 2°53 0:9 25 (0) 25 5 192 The process of filtration does not always produce an inactive residue, as on several occasions the residue after very thorough washing has been found to retain a considerable amount of activity. No reason has yet been found for this and it has not yet been ascertained whether it is due to some peculiarity in the particular specimen of juice or in the special filter employed. It is of interest to note that in Experiment 2 (Table V) the residue alone gave 8°3 ¢.c. of carbon dioxide in 3 hours, the amount evolved in the last hour being only 0-1 cc. At the close of this period the liquid still contained the alcoholic enzyme, since on the addition of 20 c.c. of the filtrate, fermentation recommenced and continued for many hours. These two sets of experiments (Tables IV and V) show that the fermentation of glucose by yeast-juice is dependent upon the presence of a dialysable substance which is not destroyed by heat. 4. Analysis of the Effect of the Addition of Boiled Juice upon the Fermentation of Glucose by Yeast-jusce. In order to compare the course of the fermentation in the presence and in the absence of boiled yeast-juice, experiments were carried out in which the rate of evolution of carbon dioxide was observed in each case throughout the whole period of activity of the juice, which, as a rule, in presence of an excess of sugar, lasts for about 48 to 60 hours. For this purpose the fermentation was allowed to proceed in a 100 cc. flask, kept at the constant temperature of 25° by immersion in a thermostat, and connected with an azotometer, in which the gas was collected over mercury. The gas in the fermentation flask was maintained at a constant pressure, as nearly as possible that of the atmosphere, by keeping the mercury 1905. | The Alcoholic Ferment of Yeast-Juice. ALI in the reservoir at a fixed level, by means of a syphon dipping into a small beaker. The volume of the gas was read on the azotometer without disturbing the mercury reservoir and was reduced to atmospheric pressure by means of a calibration curve. Since yeast-juice readily becomes supersaturated with carbon dioxide, the contents of the flask were vigorously shaken before each reading of the volume of gas. Before the observations were commenced the liquids were brought to the temperature of the thermostat, and were saturated with carbon dioxide. In all comparative experiments the concentration of glucose was the same. When the rates of evolution of carbon dioxide from (1) a solution of glucose in yeast-juice, and (2) a similar solution to which boiled and filtered yeast- juice has been added are compared, it is found that two phenomena are concerned in the production of the increased fermentation in the presence of boiled yeast-juice. (a) An initial rapid evolution of carbon dioxide is produced, which soow diminishes until a rate is attained which remains nearly constant for several hours and is usually, but not invariably, approximately equal to that given by an equal volume of the same yeast-juice and glucose to which no addition has been made. (v) The fermentation rate diminishes more slowly, so that the fermentation continues for a longer period. The greater proportion of the total increase is usually due to this second phenomenon. The results obtained in a typical experiment of this kind are shown in Fig. 1. The initial period of the evolution is plotted separately (Curves A’ and B’) on a larger scale. Curves A and A’ in which the evolution of carbon dioxide is plotted against time represent the course of a fermentation with 25 c.c. yeast-juice +25 cc. water +5 grammes glucose +toluene. The rate to begin with is 48 ¢.c. per hour, but rapidly decreases until it becomes equal to 24 c.c. per hour, at which it remains almost constant for about 5 hours, gradually decreasing until, after the expiration of about 40 to 45 hours, fermentation ceases. The total evolution amounted to 369 c.c. under atmospheric conditions. Curves B and B’ refer to 25 c.c. of the same yeast-juice +25 c.c. of a boiled yeast-juice +5 grammes glucose+toluene. The initial rate is much higher, 168 cc. per hour, but this falls gradually in the course of 40 minutes to 30 c.c. per hour. This rate of 30 cc. per hour falls off much less rapidly than that in Experiment A, the fermentation continuing for about 80 to 85 hours and yielding in all 1174 ¢.c. of carbon dioxide. It is important to bear in mind that these curves represent the gradual disappearance of the fermenting power 412 Dr. A. Harden and Mr. W. J. Young. [ Dec. 8, ea BE? SS Pf PD a pe ep cle a ea acc SIE) we A ee ate asa Oe Se ee Atcha ABER ERRRRRERSEE 0 15 20 70 75 80 8 90 95 100) ES IN HOURS. —> CARBON D/OXIDE IN C.C.8. of the liquid, and not the diminution of the amount of fermentation with diminishing concentration of sugar, an excess of this substance being present throughout. A comparison of the two curves shows very clearly the two factors involved vane cane in the great increase in Experiment B: jae tf (1) The initial rapid evolution, and (2) the prolongation of the fermenta- QO 10 20 30 40 5060 70 80 TIME IN MINUTES —> tion. 5. The Initial Period of Rapid Evolution of Carbon Dioxide. This is a very striking phenomenon, and a typical example is illustrated in fig. 2 in which the curves show the course of the evolution of carbon dioxide (total volume evolved plotted against time) during two hours in the case of: A. 25 ce. yeast-juice +75 c.c. water +10 grammes glucose + toluene. B. 25 ec. yeast-juice +50 c.c. water + 25 cc. boiled ae yeast-juice +10 grammes glucose + toluene. C. 25 c.c. yeast-juice +75 cc. boiled autolysed yeast-juice +10 grammes glucose + toluene. In B and C the initial rates are almost equal (58 ¢.c. in 10 minutes) and much greater than in A (14 c.c. in 10 minutes). In B the rate rapidly falls 1905. | The Alcoholic Ferment of Yeast-Juice. 413 off whilst in C it diminishes much more slowly. A similar initial period is also observable in A, but is not nearly so marked. 260 eal! | Pile Pete ere a pee eeeee as ESE dy ret Chota ie P ae ee : raped x baa a peor SAP CARBON D/OX/DE /N C.C.S. So) [an} 0 [O97 20> s05"40 —50'— 60° 70-80 (90° V100) “10.7 120 _ TIME 1N MINUTES —w The extra quantity of carbon dioxide evolved in this initial period may be calculated by subtracting the amount corresponding with the constant rate which is finally attained from the total amount observed. This is done graphically in fig. 2 by continuing the straight line representing the constant rate back to the axis of ordinates. The following numbers are thus obtained : for A, 166; for B, 75°4; for C, 192°9. The amounts due to the addition of boiled juice are therefore: for 1 volume in B, 75-4—16°6=58'8 ; for 3 volumes in C, 192°9—16:6=176°3 = 3 x 58'8. The extra amount of carbon dioxide is, therefore, directly proportional to the volume of boiled juice added. AlA Dr. A. Harden and Mr. W. J. Young. [ Dec. 8, 6. Production of the Initial Rapid Evolution of Carbon Dioxide by the Addition of Phosphates. As the result of a large number of attempts to isolate the constituent of boiled juice which brings about the increase in fermentation, it was found that whenever an increase was produced phosphoric acid in the form of a soluble phosphate was present. The effect of the addition of soluble phosphates to yeast-juice was, therefore, examined and it was found that a well-marked initial rapid evolution of carbon dioxide was thus produced. Since, moreover, the boiled juices employed invariably contained phosphates, precipitable by magnesia mixture, there can be no doubt that it is to the presence of these that this initial phenomenon is due. Quantitative estimations revealed the somewhat surprising fact that the extra quantity of carbon dioxide evolved in the initial period when a phosphate or a boiled juice is added, corresponds with the evolution of one molecular proportion of carbon dioxide for each atom of phosphorus added in the form of phosphate. In order to obtain accurate results with solutions of sodium or potassium phosphate, the fact that these absorb carbon dioxide must be taken into consideration. Solutions of the dihydrogen salts of potassium and sodium are too acid to be employed and the monohydrogen salts or a mixture of these with the dihydrogen salts were always used. In every case the liquid before being added to the yeast-juice was saturated with carbon dioxide at the temperature of the bath, and the volume of carbon dioxide liberated by the addition of excess of hydrochloric acid was ascertained in an aliquot portion. At the close of the fermentation the fermented liquid was acidified and the residual combined carbon dioxide measured, the difference between this and the original amount being subtracted from the amount evolved during the fermentation. The results are more precise when the yeast-juice employed is an active one, since when the fermenting power of the juice is low the initial period becomes. unduly prolonged and the calculation of the extra amount of carbon dioxide is. rendered uncertain. The equivalence of the carbon dioxide and phosphate is established by the results contained in the following Table VI. Column i gives the observed amount of extra carbon dioxide calculated as described above and reduced to grammes, and Column 2 the equivalent of the phosphate: added, this being estimated by precipitation with magnesium citrate mixture in the boiled juice or phosphate solution. In Experiments 1 to 7 boiled juice was added; in 8 to 14 a solution of sodium or potassium phosphate. The maximum rate attained during the initial period is from five to eight LS (OG ia The Alcoholic Ferment of Yeast-Juice. 415 times as high as the constant rate attained after the evolution of the carbon dioxide equivalent to the phosphate present. Table VI.—Equivalence of Extra Carbon Dioxide Evolved during the Initial Period, and Phosphate added. Grammes of carbon dioxide. | Grammes of carbon dioxide: 15;-7 560 | ganna ee] eg aD oc 3 om = ments. Column II— || ments. ap olumn II— counts 1. Calculated | Comm | Caleulated * | from phosphate. | ‘| from phosphate. 1 0 090 0 086 Suiid, sOrleG | 0-197 2 0:054: 0 055 9 | 0 066 0 :065 3 0 ‘058 0°051 | 10 | 0-057 | 0-061 4 0 ‘060 0 049 11 0 056 | 0-061 5 0-106 0-112 12 } 0 ‘059 | 0 ‘061 6 0.°103 0-101 | 13 0 ‘068 0-070 7 01138 0-112 | 14 0-071 | 0-070 At the commencement of the period when sodium or potassium phosphate solution has been added, the rate only gradually acquires its maximum value and sometimes it only attains this maximum after a considerable interval. — This phenomenon is occasionally observed in the fermentation produced by yeast-juice without the addition of phosphate, aud also sometimes occurs, but to a much smaller extent, when boiled juice is added. It is well shown in Curve B, fig. 5, which represents the fermentation produced by 25 c.c. yeast- juice +25 cc. of a 0:06 molar solution of sodium phosphate +5 grammes glucose + toluene. The cause of this period of induction has not yet been ascertained. 7. Linit of the Action of Phosphate. If the fermentation in presence of phosphate be allowed to continue until the steady rate is attained and a second quantity of phosphate be then added, a second period of rapid evolution of carbon dioxide sets in and proceeds in a similar manner to the first. This is shown in Curves B and C, fig. 3, which repre- sent the effect of the successive addition of two quantities of 5 c.c. of 0:3 molar sodium phosphate to 25 c.c. yeast-juice +20 c.c. water, in presence of 10 per cent. glucose. Curve A represents the fermentation in absence of added phosphate. The phosphate solution employed was a mixture of five molecules of NaH:PO, with one molecule of NasHPO; and no correction for combined carbon dioxide was required. The extra amount of carbon dioxide evolved after each addition is the same, and is equivalent, as VOL. LXXVIIL—B. Er 416 Dr. A. Harden and Mr. W. J. Young. [Dec. 8, already stated, to the phosphate added. The equality is shown graphically in the curve and the equivalence in Experiments 13 and 14, Table VI. CARBON O/OXIDE /N C.0.8. ——> 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 7/ME 1N MINUTES —> ' This process cannot, however, be repeated indefinitely, as after a certain limit is reached the reaction no longer occurs and with a large excess the fermentation is stopped. The exact limit appears to vary both with the nature of the phosphate added and with the particular specimen of yeast- juice employed. The greatest amount of carbon dioxide hitherto obtained in this way from 25 c.c. of yeast-juice is about 0-45 gramme (230 c.c.), which was observed on two occasions, once after the addition of four volumes of boiled juice, and again after the addition of 50 c.c. of a solution of a mixed magnesium potassium phosphate yielding with magnesia mixture 1:187 grammes of magnesium pyrophosphate. When a specimen of yeast-juice has been incubated until it will no longer ferment sugar, it is not affected by the addition of phosphate. The fact that the extra carbon dioxide calculated in this way is equivalent to the phosphate present, suggests the superposition of two actions. Whether this is to be explained by the presence of two distinct enzymes or simply by the increased activity of a single enzyme remains to be decided. 1905. | The Alcoholic Ferment of Yeast-Juce. 417 8. Products of Fermentation in the Presence of Phosphate. The carbon dioxide evolved during the initial period after the addition of a phosphate is the product of a true alcoholic fermentation of the glucose, in which alcohol and carbon dioxide are produced in equivalent amounts. This was proved in the following way. Twenty-five cubic centimetres of a solution containing 25 grammes of glucose and 5 c.c. of a 0°3 molar solution of potassium phosphate were added to 25 c.c. of yeast-juice; the mixture was incubated and the carbon dioxide collected and measured. As soon as the rate of evolution had become constant, a further addition of 10 cc. of 0°3 molar phosphate solution was made and the fermentation again continued until the rate had become constant. The gas evolved was tested and found to be carbon dioxide. The total amount evolved during the experiment, which lasted for 2 hours 10 minutes, was 163-4 cc. at 19°-6 and 758°6 mm. or 0:291 gramme, the equivalent of the phosphate added being 0-196 gramme. The liquid was then distilled with steam and the alcohol estimated in the distillate, 1312 grammes being found to be present. Twenty-five cubic centimetres of the original juice were found to contain 0:983 gramme of alcohol and therefore 1°312—0:983 = 0°329 gramme were formed by the fermentation of the sugar. The ratio of alcohol to carbon dioxide produced is therefore 0°329/0:291 = 1:13, which agrees well with the ratio previously found by similar methods for the fermentation of glucose by yeast-juice.* The theoretical ratio is 1:04. Lactic acid and acetic acid were also estimated in the original juice and after fermentation in presence of phosphate, but only a very small variation was observed. Twenty-five cubic centimetres of juice gave before fer- mentation 0122 gramme of zinc lactate and 0:083 gramme of acetic acid, and after fermentation 0°102 gramme of zinc lactate and 0:072 gramme of acetic acid. 9. Fate of the Phosphoric Acid. When the fermented liquid is boiled and filtered almost the whole of the phosphorus present is found in the filtrate, but it is nearly all in a form which is not precipitated by ammoniacal magnesium citrate mixture. In the following experiment three quantities of 25 c.c. of yeast-juice were taken :— A. Hot water was added, the solution heated in a boiling water-bath and the coagulate filtered off and well washed. B. Ten cubic centimetres of a 30 per cent. glucose solution and 10 c.c. of * Harden and Young, ‘ Ber.,’ 1904, vol. 37, p. 1052. an elees 418 Dr. A. Harden and Mr. W. J. Young. [Dec. 8, 0:3 molar potassium phosphate solution were added and the liquid at once heated to the boiling point, filtered, and the coagulate washed. C. The same additions were made as to B and the liquid then fermented until the close of the initial period, after which it was heated and filtered like the others. The total phosphorus was then estimated in each of the coagulates and in each of the filtrates, and the phosphorus precipitated by magnesium citrate in each of the three filtrates. The estimations of total phosphorus were made by heating with sulphuric and nitric acids until colourless, diluting and precipitating with magnesium citrate mixture in presence of excess of ammonia. The following were the results obtained, the numbers representing the grammes of magnesium pyrophosphate per 25 c.c. of juice. Table VII. A. B. C. SOAS etn Juice + phosphate. | Juice + phosphate. | Original juice.) Not fermented. Fermented. Coagulate ............0.. 0053 0-057 0-072 Filtrate— (a) Precipitated by Meg citrate ......... 0 °126 0-480 0:070 (6) Not precipitated by Mg citrate ... 0:271 0-282 0-679 Total -..... 0-450 0°819 0 °821 | The amount of phosphate added was equivalent to 0372 gramme of magnesium pyrophosphate. , A number of other results are given to show the extent to which phosphate is converted into the non-precipitable form by this reaction. All the estimations were made by boiling and filtering the fermented liquid immediately upon the close of the initial period. As before the numbers represent grammes of magnesium pyrophosphate obtained from 25 c.c. of juice. The form in which this non-precipitable phosphorus is actually present in the fermented liquid, and in the liquid which has been boiled and filtered, has not yet been ascertained with certainty. Experiments which are still in progress, however, appear to indicate that it exists in combination with glucose, probably in the form of a phosphoric ester. ‘ 1905. | The Alcoholic Ferment of Yeast-Juice. 419 Table VIII.—Conversion of Phosphate into the Non-precipitable Form by Yeast-juice and Glucose. Phosphate Precipitable Non-precipitable added. phosphate in filtrate. | phosphate in filtrate. 1 0553 0 ‘066 1 °032 2 0-490 0-090 0 832 3 0-250 0054 0 °685 4 0-488 0-091 1-040 5 0 °495 0-088 0-881 The question as to whether the entire phenomenon of the fermentation of glucose by yeast-juice depends on the presence of phosphates has not yet been definitely decided. The addition of phosphate undoubtedly produces a larger increase ,in the total fermentation than is simply due to the equivalent amount of carbon dioxide evolved in the initial period. The extent of this increase appears to vary very considerably with different specimens of yeast-juice, but the prolongation of the fermentation is not so great as is caused by boiled fresh juice. This question can only be satisfactorily settled by ascertaining whether the addition of a phosphate to the perfectly inactive residue obtained from a juice by filtration through a gelatin filter is sufficient to restore its fermenting power in the same way as the filtrate or a boiled juice. Experiments on this point are in progress, but no decisive result has as yet) been obtained, and all discussion of this point will best be deferred until these are completed. Various other points of interest raised in the course of the investigation, and the study of the relation of these phenomena to the fermentation of glucose by living yeast, are also occupying our attention. A short outline of the main conclusions arrived at in the foregoing paper, has been previously published in the form of two preliminary com- munications, without any experimental details.* After the appearance of these notes, Buchner and Antoni}+ repeated and contirmed a number of the experiments dealing with the effect of boiled juice and of phosphates on the total fermentation, and with the separation of the juice by dialysis into an inactive residue and a dialysate capable of rendering it active. Buchner and Antoni were able, with the more stable juice at their disposal, to carry out the dialysis in the ordinary way for 24 hours and in this manner to confirm the results obtained by the use of the gelatin filter. Owing to the * ‘Journ. Physiol.,’ 1904, vol. 32 ; ‘ Proc.,’ of November 12; ‘Proc. Chem. Soc.,’ 1905, vol. 21, p. 189, June 6. t ‘Zeit. Physiol. Chem.,’ 1905, vol. 46, p. 136. 420 Mr. Armit and Dr. Harden. Estimation of — [Dee. 5, lack of experimental detail, Buchner and Antoni imagined that in our comparative experiments the concentration of glucose and of enzyme had not been kept constant, and ascribed part of the increase produced by boiled juice to the favourable effect of a diminution in the concentration of the sugar and of the alcohol, which is always present, by dilution with the added boiled juice. The details given above show that neither of these influences had any share in the effects observed by us. The Quantitatie Estimation of Small Quantities of Nickel mn Organic Substances. By H. W. Armit and A. Harpen, D.Sc., Ph.D. (Communicated by Dr. C. J. Martin, F.R.S. Received December 5, 1905,— Read February 1, 1906.) _ (From the Chemical Department of the Lister Institute of Preventive Medicine.) In the course of an investigation into the toxic action of certain nickel compounds, it was found necessary to devise a method of detecting and estimating nickel, when included in animal tissue, in quantities not exceeding a few milligrammes per cent. A method has therefore been worked out, which, although in many respects only differing from the usual methods in virtue of slight alterations of detail, is capable of demonstrating extremely small quantities of nickel accurately. The method may be divided into three stages: (1) The Ashing ; (2) The Separating ; and (3) The Estimating stages. 1. Ashing.—The substance to be examined must be placed in a porcelain crucible (platinum is unsuitable, as a considerable loss of nickel takes place, probably by an alloy of platinum and nickel being formed) and evaporated to dryness over a water bath. If the substance be solid, it should be cut up into small pieces. The crucible is then heated carefully with a Bunsen flame, but it may be wise to further dry in a hot air oven or on a sand bath before this. Then it is burned over a Fletcher burner, and lastly fully incinerated in the blow-pipe flame. With some care, it is possible, as a rule, to oxidise fully all the carbon, without recourse to any foreign material. The crucible . is then placed on the water bath, and 10 ¢.c. of pure hydrochloric acid are added and allowed to evaporate to dryness, this process being repeated. The residue is then extracted with water to which a small quantity of hydro- 1905.] Small Quantities of Nickel in Organic Substances, AZL chloric acid is added. For this purpose, 2 c.c. of a four times normal acid are generally employed. The extract is then filtered. The ash so obtained is practically completely soluble. Ashing by Kjeldahl’s method, or better, with sulphuric and nitric acids, can also be employed, but has two disadvantages over the simple incineration method: Firstly, it takes longer; and secondly, it introduces foreign salts, which should, if possible, be avoided. 2. Separation.—Firstly, it is necessary to get rid of the iron and at the same time of the phosphates. Those tissues, which contain iron in excess, eg., blood, may be treated by precipitation with excess of ammonia and filtration. The process should be repeated three times, the precipitate being redissolved each time with the same quantity of acid as was used for the extraction. If some of the iron separates out from the filtrate on being evaporated, it may be necessary to refilter before dryness is reached. When the substance contains little or no iron, it is necessary to add a sufficiency to combine with the quantity of phosphates present. If the phosphates are present in excess, the following method is employed. The cold solution is made neutral to litmus or very faintly acid with ammonia. An excess of ammonium acetate is then added (as a rule 8 to 10 cc. of a 10-per-cent. solution suffices), and sufficient ferric chloride to colour the supernatant fluid yellowish-red. The mixture is then boiled, when all the iron separates out as phosphate and basic acetate. For those tissues yielding large quantities of phosphates, ¢.g., liver, the amount of ferric chloride necessary is comparatively large, and some difficulty may be experienced with the filtration. The only possible help is obtained by using two or more filters. The washing of the precipitate must be carried out carefully, in spite of the considerable loss of time. Precipitation by ammonia may be carried out for the three subsequent repetitions. The precipitate is each time dissolved in the smallest possible quantity of acid. After the united filtrates have been evaporated to dryness, the residue is again dissolved in water, and dilute hydrochloric acid added, about 6 c.c. of four times normal acid being usually sufficient. Sulphuretted hydrogen is then passed through the hot solution for at least half an hour, and it is then allowed to stand for a time, as the sulphides, which form in acid solution do not always readily separate out. The liquid is then filtered and the precipi- tate well washed with sulphuretted hydrogen water. The filtrate is again evaporated to dryness, re-dissolved in a little water on the water bath and then a solution of pure sodium hydrate is added in successive portions to the hot liquid, until no more ammonia comes off. Care should be taken to use as little sodium hydrate solution as possible, as every sample in the market 422 Mr. Armit and Dr. Harden. Estimation of [Deec. 5, contains small traces of iron. The nickel is thus precipitated in the form of the hydrate, and this is converted into nickel sesquioxide by the addition of 1 or 2 cc. of bromine to the cold mixture. The nickel oxide is then collected by filtration, and after having been well washed, is dissolved in hydrochloric acid, and the solution evaporated to dryness to remove the excess of acid, and the residue re-dissolved in water with a-faint trace of acid, in order to prevent the formation of basic salts. The solution is finally made up to a definite volume. In the process of separation, no especial difficulties save the management of the voluminous iron precipitate, are met with as a rule. At times an insoluble residue is found on the filter paper when the oxide is dissolved. This is a trace of a sulphide of copper or another metal of this group, which has escaped precipitation by sulphuretted hydrogen in acid solution. 5. Estimation—The usual method of quantitative estimation of nickel colorimetrically is carried out with ammonium sulphide, but it has been found that sharper results can be obtained by employing a-dimethylglyoxime CH:;.C(N.OH).C(N.OH).CHs, which was recently shown by Tschugaeff to form a scarlet red compound with nickel in the presence of ammonia.* For this purpose, a saturated solution of the reagent in absolute alcohol is prepared, this is diluted with water until a little of the compound separates out, and alcohol is then added until complete solution takes place. The fluid to be tested and a standard solution of nickel sulphate are placed in burettes. A measured quantity of the fluid is then run into a Nessler tube and to this 0°5 cc. of a 10-per-cent. solution of ammonia and the same quantity of the dimethylglyoxime solution are added and the whole made up to 30 cc. It is better first to add the ammonia to the nickel solution, then the dimethylglyoxime, and then allow the colour to develop before diluting up to the 30 cc. mark. All the solutions must be cold. The fluid becomes coloured pinkish red, the depth of the coloration depending on the quantity of nickel present. The most convenient quantity to work with is about 0:08 to 0°01 milligramme. The colour is then compared with that produced by varying quantities of nickel from the standard solution. The determination is not complete until a quantity has been found, which gives a colour which is just too pink, and a second quantity the colour of which is just appreciably less pink, than the fluid to be tested. The quantity of nickel contained is then calculated as the amount midway between the two tubes. With a little practice, it is quite easy quickly to determine very small differences of colour. The estimation should be concluded as rapidly as is compatible with accuracy, * ‘Dent. Chem. Ges. Ber.,’ 1905, vol. 38, p. 2520. 1905.| Small Quantities of Nickel in Organic Substances. 423 as, after a short time, the nickel compound with dimethylglyoxime separates out of the coloured solution as a precipitate. The advantages of this method over the ammonium sulphide method are : (1) small traces of iron do not interfere with the final colour, nor with the sharpness of the method; (2) smaller quantities of nickel can be accurately estimated ; and (3) it is easier to work in a bad light with the pink than with the brown colorimetric determination. Dealing with the colorimetric test alone, with solutions of pure nickel sulphate, the smallest quantity which gives the reaction is 1/1000 milligramme. To detect such a small amount, the solution must be placed in the Nessler tube, then the ammonia and the solution of dimethylglyoxime added, when one can recognise the characteristic pink colour, and, lastly, the fluid is made up to 30 ec.; on comparing this with distilled water, a faint but distinct difference is seen. 3/1000 milligramme gives a recognisable pink colour in 30 cc. of fluid. Working with 0:07 milligramme, differences of 1/1000 milligramme can be recognised with a little practice. This represents a potential error of + or —0°7 per cent. In test analyses, serum or blood with nickel sulphate added, the experi- mental error was kept as low as 2 per cent., using about 1 milligramme of nickel. Fer example, about 30 grammes of blood were placed in a crucible and 0:9 milligramme of nickel, in the form of the dissolved sulphate, was added. After ashing, extracting, removing the copper and iron groups, and precipitating the nickel in the form of the sesquioxide, etc., the final fluid was made up to 30 cc.; 2 cc. of this fluid were compared with varying quantities of a solution containing 0:01 milligramme of nickel per cubic centimetre. It was found that 59 @.c. gave a colour, which was just too pink, and 5°8 c.c., a colour, which was not pink enough, so that the 2 cc. contained 0:0585 milligramme of nickel, and the whole solution 0°88 milligramme. This represents a loss of 0:02 in 0°9, or about 2 per cent. Electrolysis is only to be preferred when large quantities of nickel are to be measured, while the method described above is intended for the recognition and measuring of quantities of nickel not exceeding a few milligrammes. On Voges and Proskauer’s Reaction for Certain Bacteria. By Arruur HARDEN, D.Sc., Ph.D. (Communicated by Dr. C. J. Martin, F.R.S. Received December 5, 1905,— Read February 1, 1906.) (From the Chemical Laboratory Lister Institute.) In 1898 Voges and Proskauer* described a new colour reaction which they had observed in the case of a bacillus, isolated by Voges and grown in a medium containing sugar. When potash was added and the tube allowed to stand for 24 hours or longer at room temperature, a beautiful fluorescent colour, somewhat similar to that of a dilute alcoholic solution of eosin, formed in the culture fluid, particularly at the open end of the tube exposed to the air. The reaction was found to be specific to the bacillus in question, and was not given by any of the other organisms isolated in the course of the investigation upon which they were engaged, nor by the B. colt communis, so that it afforded a most valuable means of differentiation for the inhabitants of the intestine. Durham} and How} have also employed this reaction for the discrimination of intestinal bacteria, and MacConkey,§ in confirmation of Durham, has found that out of a large number of bacteria which were tested only three gave the reaction, these being B. lactis aerogenes (Kscherich), B. capsulatus (Pfeiffer), and B. cloacw (Jordan). The examination of the products formed by &. lactis aerogenes from glucosel| has shown that acetylmethylearbinol, CH;.CO.CH(OH).CHs, and 2°3-butyleneglycol, CHs3.CH(OH).CH(OH).CHs, are both present in the medium in which this has been cultivated in the presence of glucose. The acetylmethylcarbinol has not as yet been isolated in the pure state, but is present in the aqueous distillate obtained by distilling the culture medium. This distillate and the glycol were, therefore, treated with caustic potash in order to ascertain whether either of them was the cause of the reaction just described. Neither of these substances produces the characteristic _ fluorescent coloration with potash alone, but when peptone water is also added, acetylmethylearbinol gives the reaction after standing for about 24 hours, whilst the glycol does not react in this way even on standing. The coloration was produced in the characteristic manner described by ‘Zeitschr. f. Hyg.,’ 1898, vol. 28, p. 20. ‘Journ. of Experimental Medicine,’ 1900—1901, p. 354. ‘Centralbl. f. Bakter.,’ 1904, vol. 36, p. 484. ‘Journ. of Hyg.,’ 1905, vol. 5, 349, || Harden and Walpole. Hot+ + On Voges and Proskauer’s Reaction for Certain Bacteria. 425 Voges and Proskauer, commencing at the open end of the tube exposed to the air. This suggests oxidation as a factor in the phenomenon, and as acetylmethylearbinol is very readily converted by oxidation into diacetyl, CH;.CO.CO.CHs, this substance was tested. Diacetyl yields the fluorescent red coloration with peptone water and caustic potash in a few minutes, and by its aid a much greater depth of colour can be obtained than that observed with bacterial cultures. Voges and Proskauer’s reaction, therefore, appears to be due to acetylmethylearbinol, which is formed by the action of the bacteria on the glucose of the medium. In the presence of potash and air this is oxidised to diacetyl, which then reacts with some constituent of the peptone water. That diacetyl is the active substance and not p-xyloquinone, CsH202(CHs)», which is readily formed from it by the action of alkalis, is shown by the fact that if the diacetyl be allowed to stand for some time with potash solution, and peptone water be then added, no reaction occurs. B. cloace (Jordan), which gives Voges and Proskauer’s reaction, was also found to yield acetylmethylearbinol, which was recognised by its power of reducing Fehling’s solution in the cold and of yielding the characteristic phenylosazone of diacetyl with phenylhydrazine. Acetylmethylcarbinol has also been observed as a product of the action of certain other bacteria on glucose. Thus, Grimbert* found that it is produced by £. tartricus, and Desmotsft that it is also formed by the various bacilli of the mesentericus group and by B. subtilis and Tyrothrix tenuis, These bacteria should, therefore, give Voges and Proskauer’s reaction, and, as a matter of fact, B. mesentericus fuscus, the only one which has so far been examined, gives the reaction quite characteristically when grown in peptone water containing 2 per cent. of glucose. A number of other bacteria are being examined and attempts are also being made to ascertain what constituent of the peptone water it is that reacts with the diacetyl. * ‘Compt. Rend.,’ 1901, vol. 132, p. 706. + ‘Compt. Rend.,’ 1904, vol. 138, p. 581. 426 A Further Communication on the Specificity and Action in Vitro of Gastrotoxin. By Cuarues Bouton, M.D., Research Scholar of the Grocers’ Company. (Communicated by Professor Sidney Martin, F.R.S. Received January 25,— Read February, 1, 1906.) (From the Pathological Laboratory, University College, London.) [Puates 16 anp 17.] In July, 1904, I laid before the Royal Society a preliminary communica- tion on the production of a gastrotoxic serum. In that communication I stated that the serum, obtained by the injection of the mucous membrane of the stomach of the guinea-pig into the rabbit, was not speeific in the true sense of the word ; and further, that I was unable to demonstrate any effect of the gastrotoxin upon the gastric glands of the guinea-pig in viro, although necrosis and ulceration of the mucous membrane of the stomach were produced by injection of the serum into the living animal. By means of more extensive experiments and improved methods I have obtained confirmatory evidence that the action of the serum is not truly specific, and have also been able to demonstrate that a definite effect upon the gastric cells is produced in vitro. I have also succeeded in preparing a gastrotoxic serum by injecting the fresh mucous membrane of human stomach into the rabbit. The present communication is therefore intended as a continuation of my former one. The subject will be discussed under the following headings :— I. Action ta Vitro— 1. Hemolytic action. . Action upon the gastric granules (agglutination). . Action upon the soluble proteids of the cells (precipitation). 4. Action upon the intact gastric cells (lysis). ‘oo NOD If. SPECIFICITY OF GASTROTOXIN— 1. Power of Different Celis to Render the Serum Inactive. Experiments 77 vivo. Experiments 7 vitro. 2. Comparison with Entero- and Hepatotozin and Hemolysin. Experiments 72 vivo. Experiments ia vitro. III. Propuction or HuMAN GaAsTROTOXIN. ITV. GENERAL CONCLUSIONS. On the Specificity and Action in Vitro of Gastrotoxin. 427 I. ActTION in Vitro. In my first communication I stated that the blood serum of the rabbit injected with stomach cells washed free from blood became more highly hemolytic for guinea-pig’s red-blood corpuscles than it was previous to injection. I have recently been able to demonstrate that two distinct heemolysins are produced during the process of immunisation. The gastrotoxin has likewise the power of producing marked changes in the soluble proteids, and also in the protoplasmic granules of the gastric cells. It further brings about slight though definite changes in the intact cells themselves. 1. Hemolytic Action. Method.—In obtaining the guinea-pig’s blood corpuscles to test the emolytie power of the immunised rabbit’s serum, the blood is whipped and ccentrifugalised, the serum then pipetted off and the corpuscles washed in 0°86-per-cent. salt solution several times. A 5-per-cent. suspension of the corpuscles in salt solution is used for testing the serum. A known quantity, usually 1 c.c., of this suspension of corpuscles is mixed with diminishing amounts of the serum in a series of test-tubes, the volume of fluid in each tube being made up to the same amount with salt solution. The tubes are placed in the incubator for one hour, and then in the ice chamber till the following morning, when the exact point at which the corpuscles are completely dissolved can be determined with ease. As an example, the details of the following experiment which was made to determine the normal hemolytic power of a rabbit’s blood for guinea-pig’s corpuscles are given :— oe Normal rabbit’s Cale colation: After 1 nore sea and ice suspension). serum. chamber. C.C. ce: c.c. 1 2°75 0°25 Complete solution. 1 2°5 0-5 a 1 2°25 0°75 * 1 2 1 i | 1 1-75 1-25 Almost complete. 1 1°5 1-5 Incomplete. 1 1°25 1°75 es | 1 1 2 e 1 0°75 2-25 i 1 0-5 25 i : 0°25 2 ue Fluid above corpuscles uniformly . Ol 2 tinted. 1 Bede 075 2-25 | 1 ties. 1 O08 2-5 Fluid tinted to diminishing heights : ~ (0-025 2°75 | above corpuscles. - 0-01 2 1 coat 0075 ioe No laking: 1 - 0-005 2°53 times. 2 1 0 0025 2°75 7 428 Dr. C. Bolton. On the [Jan. 25, In the test-tube containing 2 cc. serum there was complete solution of the corpuscles, and in the test-tube containing 0°01 c.c. serum there was a trace of hemoglobin diffused in the clear fluid just above the deposited corpuscles. From the experiment it was therefore found that 2 e.c. serum of this rabbit would completely lake 1 c.c. of a 5-per-cent. suspension of guinea- pig’s corpuscles, and that a dilution of the serum of 1 in 400 was the greatest which would produce any solution at all. It is not reliable to estimate the hemolytic power solely by finding the ereatest dilution in which any solution will occur, or in other words the vanishing point of hemolysis, because this vanishing point may occur in higher dilutions in the case of a weaker serum than in the case of a stronger serum. What is the exact reason for this phenomenon does not appear to be at all clear. Gay (1) has, however, recently shown that in the case of high dilutions the activity of the complement may be completely inhibited. Bashford (2) has suggested that in the higher dilutions hemolysis is interfered with by agglutination of the red corpuscles. The hemolytic power of most rabbits is fairly constant, but as they vary somewhat within small limits, I examine the hemolytic power of each rabbit’s blood before injection. Care must be taken to use exactly the same dilutions when the serum is subsequently tested, as the amount of dilution affects the heemolytic power of the serum. Hemolysis—A few days after the first injection of guinea-pig’s stomach cells into the peritoneal cavity of the rabbit, the hemolytic power of the rabbit’s serum for guinea-pig’s red corpuscles is found to have considerably increased. In one case before injection 2°25 c.c. serum were necessary to completely dissolve 1 ¢.c. of a 5-per-cent. suspension of corpuscles; seven days after the injection 0°75 c.c. serum would dissolve the same amount of corpuscles. This first increase of hemolysin is a true increase of the natural hemolysin of the rabbit, because its action like that of the natural hemolysin is destroyed by heat, and is not restored on adding normal guinea-pig’s serum. In other words, guinea-pig’s complement will reactivate neither. At this early stage a slight amount of laking, varying in extent in different cases, may be seen on reactivating the heated serum with guinea-pig’s complement, but this laking has never been to any degree extensive enough to account for the increase of hemolysis. In the above experiment, 21 days after the first injection, it was found that 0°d c.c. serum would completely. dissolve 1 c¢.c. suspension of corpuscles, but on heating 2°5 c.c. immune serum to 55° C., and complementing with 0°25 c.c. guinea-pig’s serum, only 1906.| Specificity and Action in Vitro of Gastrotoxin. 429 the faintest trace of hemoglobin was seen to be diffused in the fluid immediately above the deposited corpuscles. After the second injection, however, it is found that guinea-pig’s serum will reactivate the heated immune serum to a considerable extent. In the above experiment six days after the second injection 0°5 c.c. heated immune serum, on being reactivated by 0:25 cc. guinea-pig’s serum, completely dissolved the test amount of corpuscles. I have confirmed this experiment several times, and to my mind it conclusively points to the presence of two distinct immune bodies. (1) An increase of the normal hemolysin of the rabbit which 7s not comple- mented by guinea-pig’s serum. (2) A newly-formed and therefore artificial hemolysin which zs complemented by guinea-pig’s serum. This has an important bearing upon the hypothesis of the multiplicity of immune bodies, which is upheld by Ehrlich and Morgenroth (3), but is denied by other observers, notably Muir and Browning(4), and Gay (5) working in the Pasteur Institute. Agelutination of the red corpuscles also occurs. 2. Action upon the Protoplasmic Granules of the Gastric Cells. Method.—The method which I employ in order to demonstrate this action is an imitation of that described for hemolysin. The mucous membrane of a guinea-pig’s stomach is first washed free from blood by sterilised salt solution, which is made to flow through a canula introduced into the thoracic aorta, the stream issuing from the inferior vena cava. It is then scraped off, and the pulp ground up in a glass mortar. An emulsion is made with salt solution and centrifugalised for five minutes at alow speed. The supernatant fluid on being pipetted off is found to contain in suspension innumerable large and small protoplasmic granules. The granules are separated from the albuminous fluid in which they float by centrifugalisation at a high speed, and repeatedly washing until the washings give no precipitate with potassium ferrocyanide and acetic acid. The final suspension of granules in saline solution, which is to be used, must be well agitated so as to free all the granules, and before use it must be slowly centrifugalised to ultimately remove any masses of granules or pieces of tissue which happen to be present. A series of test-tubes is prepared, each tube containing 2 c.c. of the gastro- toxic serum in increasing dilutions. The first tube contains undiluted serum, and the remainder dilutions from 1 in 5 tol in 320. To the contents of each tube three to five drops, according to concentration of the suspension of granules in salt solution, are added. A control of normal saline is prepared, and also a control in which 430 Dr. C. Bolton. On the | [Jan. 25, 0:5 c.c. of the washings is added to 2 cc. serum, so as to eliminate any possibility of a precipitate of albumin being thrown down from the fluid, in which the granules are suspended, by the immune serum. The tubes are incubated for four or five hours and then examined by the naked eye and also microscopically. Agqglutination.—A very fine deposit consisting of agglutinated granules is seen, sometimes only with a lens, at the bottom of the tubes up to a certain dilution which varies according to the strength of the serum. The super- natant fluid contains granules in all stages of agelutination, and if the tubes are allowed to stand in the ice chamber till the next day, all the agglutinated granules are found to have settled to the bottom. The reading of the tubes is taken at the end of four or five hours’ ineuba- tion, and controlled by that taken on the following day, both macro- and microscopically. The reason for taking the reading twice is because bacteria are liable to be found in the fluid at the end of 24 hours, since the scraped mucous membrane cannot be sterilised by heat. The saline control shows no such deposit or agglutination, and the control containing the washings shows no precipitate or deposit. The blood of a normal rabbit shows no deposit and no agglutination, and can therefore be used as a control. The amount of deposit is just sufficient to be accounted for by the sub- sidence of the granules, and on shaking it up the agglutinated granules pass into suspension, forming a delicate precipitate. Vigorous shaking disentangles the granules, and the appearance of the solution becomes the same as it was when the granules were first added. ¢ is thus evident that the serum possesses an action upon the granules themselves, and that this action is similar to that of the bacterial agglutinins. The agglutinin appears in the rabbit’s blood in small amounts about 14 days after the first injection, and can be quite easily recognised after a second injection. Lffects of Heat——After exposure to a temperature of 58° to 60° for half an hour the serum agelutinates gastric granules in as high a dilution as it does when unheated. In this property of resisting heat the agglutinin resembles those of bacterial origin. 3. Action upon the Soluble Proteids of the Gastric Cells. Method.—The solution of proteid is prepared by grinding up the mucous membrane of the guinea-pig’s stomach, previously washed free from blood, as described above, in a glass mortar with normal salt solution. The emulsion is centrifugalised in order to get rid of most of the solid 1906.| Specificity and Action in Vitro of G'astrotoxin. 431 matter, but it is not possible to completely free such a solution from protoplasmic granules by cenrtifugalisation alone. The solution is filtered through a Berkefeld filter. The resulting filtrate, which is a perfectly clear solution like water, gives a precipitate with heat, ferrocyanide of potassium, and acetic acid, and other precipitants of proteids. The experiment is done in the same way as that described for the agglutination test. The gastrotoxic serum is diluted and placed in a series of test-tubes, and to the contents of each test-tube 0°5 c.c., or even only two or three drops of the proteid solution, are added. The tubes are placed in the incubator for four or five hours. Controls of normal rabbit’s serum and salt solution are also prepared. Precipitation.—At the end of one hour’s incubation a fine precipitate has commenced to form in the solution. This precipitate becomes coarser in appearance, and at the end of about four hours has settled to the bottom of the tube in considerable quantity, forming a deposit. Flakes of precipitate ean still be seen floating about in the otherwise perfectly clear supernatant fluid. Under the microscope this precipitate is seen to consist of amorphous- looking masses. If the tubes are placed in the ice-box until the next day the whole of the precipitate will be found to have settled to the bottom, leaving the supernatant fluid perfectly clear. As in the case of the agglutination test, I record the result of the experiment both at the end of four or five hours, and also on the following day, although in this case one can exclude the presence of bacteria, since the tubes have been sterilised and plugged with cotton wool, and the gastric solution filtered. The control tube containing normal rabbit's serum, and that containing salt solution, show no precipitate after incubation. It is thus perfectly clear that the gastrotoxic serum acts chemically upon the soluble proteids of the guinea-pig’s gastric cells, producing an insoluble compound. This precipitin appears in the serum about the same time as the agglutinin. Effects of Heat—As in the ease of the agglutinin, exposure to a tem- perature of 58° to 60° C. for half an hour does not in the least diminish the power of the serum to precipitate the soluble proteid, since precipi- tation occurs in as high a dilution of the heated serum as it does in the unheated serum. It is well known that the effects of heat upon the action of precipitins varies considerably ; some are easily affected, whilst others are resistant. Action upon Guinea-Pig’s Blood Serum—The experiment is performed in exactly the same manner as described above. The guinea-pig’s serum is ‘diluted 10 or more times, and 0°5 c.c. added to 2 c.c. gastrotoxic serum. VOL, LXXVII.—B. 21 432 Dr. C. Bolton. On the [Jan. 25, The mixture is incubated, and at the end of four hours it is seen that a similar precipitate has formed. The gastrotoxic serum may precipitate the serum in as high a dilution as it does the proteid of the gastric cells. The gastrotoxin does not therefore act exclusively upon the proteid of the gastric cells. +, Action upon the Intact Gastrie Cells. Method—aA_ guinea-pig’s stomach is washed quite free from blood before removal from the body, as described above. The superficial portion of the mucous membrane is gently scraped off with a knife and suspended in a few cubic centimetres salt solution. The test-tube is now very carefully shaken up for a few seconds and allowed to stand for about 10 minutes, at the expiration of which time the contents will have separated into two portions :— 1. A milky fluid. 2. Small pieces of mucous membrane, which either float on the top of the fluid or settle to the bottom. The milky fluid is the portion used. It is pipetted off, and on microscopic examination is found to contain in suspension free cells, masses of cells, and fragments of glands, together with broken-up cells, free nuclei, and proto- plasmic granules. It is quite easy to separate the cells, because on slow centrifulgalisation the cells and fragments of glands sink to the bottom, but the granules and nuclei remain in suspension. The cells are washed to clear away the granules and soluble proteid. A suspension of cells, masses of cells, and fragments of glands in salt solution is thus obtained. The free oxyntic cells are large oval structures with well-defined granules, but the central cells, which are of a more delicate structure and contain finer granules, tend to cling together in masses and are easily broken up if too vigorous shaking is employed. In doing an experiment three or four drops of suspension of cells are placed in 2 c.c. gastrotoxic serum in a test-tube, and the mixture incubated for four or five hours. Controls are prepared with normal rabbit’s serum and also salt solution. Lysis—The deposit which has formed in the tube is examined micro- scopically. I have done a large number of these experiments, and always with the same results. I have never observed either solution or agglutination of the cells, such as has been described in the case of other cytotoxins; in fact, the cells in the serum are much more ‘separated than 1906.| Specificity and Action in Vitro of Gastrotoxin. 433 those in the salt solution, as the latter tend to stick together by reason of the mucus, which it is impossible to clear away completely. The cells in the salt solution are quite normal in appearance, but the cells which have been exposed to the action of the gastrotoxic serum have become more or less hyaline. The oxyntic cells are not so much affected as the central cells, the masses of which appear like pieces of floating glass. The granules are obscured, and many of the cells look like shadows. The nuclei can usually be seen except when the cells are massed together. The cells are examined in the fresh state, as this seems to me to be the best for practical purposes. I have also stained them with watery solutions of methyl green, picrocarmine, and safranin, but this does not materially assist. The stained cells which are affected look like pieces of coloured glass. This effect of the gastrotoxic serum appears later than those described above, and is not observed until after the expiration of about five weeks after the first injection, four or five injections having been given in the meanwhile. The effect begins to pass off between three and four months after the first injection, the animal receiving injections at regular intervals. The serum of a normal rabbit differs in no respect in its behaviour towards the gastric cells trom salt solution. Effects of Heat—A temperature of 55° C. maintained for half an hour does not destroy the action of the serum upon the gastric cells. It may be weakened, but only to a slight degree. This result points to the conclusion that a complement exists in the cells themselves, if the action is due toa cytolysin of the same construction as a hemolysin. Such endocellular complements have of course been described before. I have heated the stomach to 54° C. for half an hour before obtaining the cells, and have found that in this case the serum produces no effect upon them. This experiment of course proves nothing, because it is probable that the vitality of the cells is destroyed by exposure to this temperature, since cell globulin coagulates at 48° to 50° C. Removal of Gastrolytic Factor—By saturating the serum with washed gastric cells of the guinea-pig and allowing the mixture to stand for one hour, the gastrolytic factor is removed and the action of the serum upon the cells is destroyed. No change is visible in the cells which have been used to saturate the serum in the space of one hour, and, since they remove the gastrolysin, it is evident that the latter becomes anchored to the gastric cells preparatory to acting upon them. This brings the gastrolysin into line with a hemolysin in this respect. 212 434 Dr. C. Bolton. On the [Jan. 25, Il. SPECIFICITY OF GASTROTOXIN. If this serum be strictly specific for the stomach cells of the guinea-pig, it should possess chemical affinities for the stomach cells alone and for no other tissues of the body. On injection into the guinea-pig it would then produce lesions limited to the stomach, and would also be rendered inactive by mixture with stomach cells, which would combine with the poisonous substance in the serum and take it out of solution. Further, other tissues of the body should not possess chemical affinities corresponding to those of the stomach cells and, therefore, would not render the serum inactive by mixture with it, and sera formed against those tissues would not produce lesions in the stomach on injection. The specificity has therefore been tested by comparing the relative power of different cells of the body of the guinea-pig to render the serum inactive, and also by comparing the effects of the serum with those of a hepatotoxic, an enterotoxic, and a hemolytic serum obtained by injecting blood. The experiments have been conducted 2 vivo and in vitro. 1. Power of Different Cells to Render the Serum Inactive. In my former communication I gave the results of a few experiments. I have since examined this question more extensively and although the subject is far from being completed yet I will here give the results of my further investigations. In mixing the various cells with the serum to be examined, care must be taken that enough cells are present to saturate the serum otherwise when they settle to the bottom of the tube a portion of the serum is left unexposed to their action. The cells obtained by scraping the mucous membrane of one guinea-pig’s stomach are enough to saturate 4 c.c. serum, but the resulting mixture will not yield 4 c.c. serum back. For a reliable experiment it is necessary to inject at least 10 c.c. serum and, therefore, to obtain this I take 12 cc. serum and mix with it the mucous membrane of three guinea-pigs’ stomachs. After centrifugalisation it is quite easy to obtain 10 c.c. of treated serum. The cells of the mucous membrane of the small intestine are obtained in the same way as those of the stomach. The liver is pounded up and passed through a tea strainer in order to prepare it. In all cases the blood is washed out of the organ in question before it is removed from the body. If the serum is examined 7m vivo after such treatment a control animal is, 1906.| Specaficity and Action in Vitro of Gastrotoxn. 435 in each case, injected with the untreated serum; if it is examined in vitro controls are likewise prepared in all cases. All the experiments have invariably been carried through to a finish on the same day, and all the tissues and sera used were always obtained fresh on that day. In testing the hemolytic power of the treated serum equal weights of the various organs were previously mixed with equal volumes of the serum. ‘The cells were allowed to stand in contact with the serum for one hour at laboratory temperature, except in the case of some of the in vitro experiments, when the tubes were placed in the incubator. Experiments in Vivo. Mixture with Stomach Cells.—Four experiments have been done, and in each case with the same result. The gastrotoxin was in all the cases completely removed by the stomach cells, and the resulting serum pro- duced no lesions in the stomach at all. The control animals in all cases showed the usual lesions (see Plate 16, fig. 1). Mixture with Intestine Cells—In four cases the intestine cells failed to destroy the action of the serum, the lesions produced by the treated serum being as extensive as those produced by the untreated serum (see Plate 16, fig. 2). In three cases the action was destroyed, but the lesions in the three control animals were so slight that the toxicity of the serum must have been very low. Mixture with Liver Cells—In four cases the liver cells failed to destroy the action of the serum, but in each case there was a weakening of the power of the gastrotoxin, judging by a comparison with the effects produced in the control animals (see Plate 16, fig. 3). In one case complete removal of the gastrotoxin resulted, and in this case the latter was of low toxic value, judging by the lesions produced in the control animal (see Plate 16, fig. 4), Mixture with Red Blood Corpuscles—In three experiments the action of the gastrotoxin was unaffected, and in a fourth the lesion was less extensive than that in the control animal (see Plate 17, fig. 5). The results of these experiments clearly demonstrate that other organs of the body besides the stomach have tissue affinities for, at all events, some of the constituents of this serum, and that they can, if not invariably destroy its action, at any rate weaken it. On the other hand, the stomach is the only organ of the body which can invariably and with uniform certainty destroy the action of the gastrotoxin. The serum is thus not, strictly speaking, specific, although lesions are not produced in other organs than the stomach by it as a rule. It may be, however, that one of the constituents of this complex serum is specific for the stomach to a great extent. 436 Dr. C. Bolton. On the [Jan. 25, These experiments also illustrate the importance of an organ’s ability to take up a poison and render it inactive without being itself affected by it. They likewise explain why large doses of the serum are necessary to produce the stomach lesions, since a large part of this serum must be rendered inactive by different organs of the body. Experiments in Vitro. Hemolysin. Mixture with Stomach Cells of Guinea- pig.—After treatment with stomach cells the serum shows as high a degree of hemolytic power as it did before such treatment. It may be higher. At first sight this result appears to be remarkable, namely, that a cell will not remove a side chain that is thrown off in response to its injection. In other words, that side chains may be thrown off which have no affinity at all for the cells against which they are thrown off. I have obtained further evidence of the same principle in the case of the gastrotoxin formed against guinea-pig’s stomach cells by injection of the rabbit’s stomach cells into the rabbit. The rabbit’s stomach cells will not remove this gastrotoxin from the serum, and therefore, whether one supposes that the cytophilic affinity of the amboceptor for the rabbit’s stomach cell is or is not saturated by an anti-immune body, the fact remains that the rabbit’s stomach cell has no affinity for the side chain which is active against the guinea-pig’s cell, and which has been thrown off in response to injection of rabbit’s stomach cell. Similarly rabbit’s stomach cells will not remove the gastrotoxin from guinea-pig-rabbit gastrotoxic serum (see Plate 17, fig. 6). It seems to me most likely that when a cell is absorbed side chains having an especial affinity for that cell, and which are used in destroying it, are set free and that other side chains having less affinity for it are set free in diminished amount, and also side chains having no affinity whatever for it are set free in smallest amount. In other words, the absorbing cell throws off most of the varieties of side chains or chemical affinities of which it is possessed, the number of each being directly determined by the amount of stimulation given to the particular chemical affinity involved. The fact that the stomach cells will not absorb the hemolysin is important from another point of view. I shall show later that the action of a hemolysin, whatever its origin, is directed especially against the stomach, and also that lesions due to its action may be limited to the stomach. Now if other organs of the body have the power of destroying the hemolysin without themselves being affected, whilst the stomach cells will not absorb it, the result naturally follows that the hemolysin will be free to act as it may in the capillaries of the stomach, and therefore will produce lesions. Mixture with Liver Cells, Intestine Cells, and Red-Blood Corpuscles——Kach of these three varieties of cells has the power of removing the hemolysin from 1906.| Specificity and Action in Vitro of Gastrotoxin. 437 the serum, and therefore of rendering it incapable of dissolving the red-blood corpuscles of the guinea-pig in vitro. Occasionally, especially in the case of the liver, a slight amount of diffused hemoglobin may be seen above the settled corpuscles. These experiments clearly indicate that although the hemolytic factor of the gastrotoxic serum may be of great importance in assisting to produce the stomach lesions in vivo, yet it 1s not the only one. The reasons for this state- ment are, that previous mixture with stomach cells will deprive the serum of its action i vivo, but will not prevent its laking blood corpuscles in vitro ; that previous mixture with liver cells, intestine cells, or red-blood corpuscles, although it deprives the serum of its power to lake blood corpuscles in vitro, will not with any degree of uniform certainty completely prevent its action in vwo. Lysin.—I have not yet attempted to compare the gastrolytic strengths of two sera by determining the highest dilutions in which any action is apparent, and therefore cannot say whether or not the action in any given case is diminished. So far as my results go, however, they appear to indicate that, after exposure of the serum to gastric cells, its action upon such cells is destroyed, but that after exposure to liver and intestine cells and blood corpuscles the serum still produces changes in gastric cells in vitro. These in vitro experiments, so far as they go, point to the same conclusion as the in vivo experiments—namely, that this gastric cytotoxin is not truly specific, although one or more of the bodies contained in it may be so, and that the protoplasmic poisons constituting it have a greater affinity for gastric cells than for the cells of other organs of the body. 2. Comparison with Hepato- and Enterotoxie Sera and Hemolysin. The hepatotoxin and enterotoxin were respectively prepared by injecting the rabbit with the washed and prepared cells of the liver and intestine of the guinea-pig. The hemolysin was obtained by injecting red-blood corpuscles, Experiments in Vivo.—Each of these sera produces hemolytic lesions in the stomach, leading to destruction of the mucous membrane, the microscopic condition very closely simulating that due to gastrotoxin (see Plate 17, fig. 7). They are more uncertain in their action upon the stomach, however, and this action is liable to be not so strictly limited to the stomach as that of gastrotoxin. The action of hemolytic serum was described in my previous communication. Experiments in Vitro. Hemolytic Power.—Both hepato- and enterotoxin 438 Dr. C. Bolton. On the [Jan. 25, possess the power of dissolving red-blood corpuscles to about the same degree as gastrotoxin. They also agglutinate these cells. Agglutination and Precipitation—The gastrotoxic serum produces similar effects upon emulsions of liver and intestine granules to those described in the case of emulsions of stomach granules. Hepato- and enterotoxin not only act upon emulsions of liver and intestine granules, but they also act upon emulsions of stomach granules. Hemolysin obtained by injecting blood has no more power of acting upon these emulsions than normal rabbit’s blood has. Lysin—Up to the present no definite hyaline transformation of gastric cells has been demonstrated as the result of the action of hepato- or enterotoxin or hemolysin, and gastrotoxin does not appear to act upon the intact liver or intestine cells. Removal of the Immune Body by Different Cells——Only one or two experi- ments of this nature have been done up to the present. In the case of a rabbit which had been immunised against the red-blood corpuscles of the guinea-pig, mixture with stomach cells entirely failed to remove the hemolysin and stomach lesions resulted on the injection of the serum (see Plate 17, fig. 8). Mixture with liver and intestine cells, however, rendered the same serum less powerful than before. In the case of an enterotoxic rabbit, mixture of the serum with either stomach cells or intestine cells effected complete removal of the amboceptor. In vitro, stomach cells completely fail to remove the hemolytic factor from enterotoxic or hepatotoxic serum, as was previously observed in the case of gastrotoxic serum. liver and intestine cells remove the hemolytic factor from entero- and hepatotoxin as they do in the case of the gastrotoxin. The few experiments that have been made with hepato- and entero- toxin therefore confirm the view that gastrotoxin is not strictly specific. III. Propuction oF HUMAN GASTROTOXIN. Nine rabbits have been immunised against fresh human stomach mucous membrane. Four died from septic infection, the remaining five gave positive results. This is not quite such an easy matter as in the case of the guinea-pig’s stomach, because the supply of human stomach is not constant, and three weeks or a month may elapse without an opportunity for obtaining the mucous membrane offering itself. In addition to this the stomach cannot be used immediately after death, although on one or two occasions I have been fortunate enough to obtain some from operation cases, and it is impossible to obtain the stomach free from blood. 1906.| Specificity and Action in Vitro of Gastrotoxin. 439 The normal serum of the rabbit is slightly hemolytic for human blood corpuscles, but does not produce any effect upon emulsions of human stomach granules, I have succeeded in showing that the sera of five rabbits so immunised became highly hemolytic for human blood corpuscles, and in the one case in which I tried it, solution of the corpuscles of the monkey also occurred. The sera also agglutinated and precipitated emulsions of human gastric granules, and in one case those of the monkey also. Whether hyaline changes are produced in the cells I have not yet determined. TV. GENERAL CONCLUSIONS. The gastric cytotoxin formed in the blood of an animal in response to the injection of gastric cells thus appears to be a complex body. After a single injection there is a great increase in the hemolysin normally occurring in the animal’s blood, and at the same time there is found a new hemolytic immune body which is not normally present in the animal. The latter is present in considerable amount after the second injection. The gastrotoxin also agglutinates red-blood corpuscles. Closely associated with the appear- ance of this artificial hemolytic immune body is that of an agglutinin which acts upon the gastric granules, and also that of a precipitin which acts upon the soluble proteids of the gastric cells. By repeating the injections these substances are found to be present in the blood for several months. Whether they are one and the same or distinct bodies I have not yet proved. After several injections, and not less than about five weeks from the first, a further substance appears in the blood, which possesses an action upon the intact gastric cells. In spite of repeated injections, this substance disappears from the blood in about four months. It is probably of the same nature as a hemolysin, but this point requires proof. The hemolytic factor is only active against blood. The actions of the agglutinin and precipitin are not confined to the constituents of the gastric cells, but extend to other proteids of the body. Whether there are separate agglutinins and precipitins for different proteids, or whether the same substances act upon all proteids, has not been determined; at all events, if the same bodies are concerned in all cases, their action upon the proteids of the stomach cells is probably greater than that upon other proteids. Whether the gastrolysin itself is truly specific remains to be proved. The few experiments that have been undertaken in the case of the human stomach indicate that the human gastric cytotoxin is identical in constitution with that of the lower animals. A40 Dr. C. Bolton. On the [Jan. 25, Note-—The term “ Hemolytic lesions” is used in this paper to signify the hemorrhages which are produced by the injection of a hemolytic serum ; this does not however imply that such hemorrhages are directly caused by the factor in the serum which brings about solution of the red blood corpuscles. REFERENCES. 1. Gay, “Observations on the Single Nature of H:emolytic Immune Bodies, and on the Existence of so-called ‘Complementoids,’” ‘Centralbl. f. Bakt., etc,’ 1 Abt., Originale, vol. 39, heft 2, 1905, p. 172. Bashford, ‘ Journal of Hygiene,’ vol. 4, 1904, No. 1, p. 40. Ehrlich and Morgenroth, ‘ Berl. Klin. Wehschr.,’ 1901, p. 569—598, and 1900, p. 681. Muir and Browning, ‘ Roy. Soc. Proc.,’ vol. 74, 1904, p. 298. Gay, loc. cit. or > go p90 DESCRIPTION OF PLATES. PuateE 16. Fic. 1.—Illustrates the removal of the gastrotoxin from the serum by treatment with guinea-pig’s stomach cells previous to its injection. Upper Stomach.—From a guinea-pig injected with gastrotoxic serum. Necrosis of the mucous membrane has therefore resulted. Lower Stomach.—From a guinea-pig injected with the same dose of the same serum previously treated with stomach cells. No lesion has resulted ; the stomach cells have removed the gastrotoxin from the serum by combining with it. Fie. 2.—Stomach of a guinea-pig which was injected with gastrotoxic serum previously treated with guinea-pig’s small intestine cells. The gastrotoxin has not been removed from the serum by the cells, and the stomach therefore shows necrosis of the mucous membrane. Fic. 3.—Stomach of a guinea-pig which was injected with gastrotoxic serum previously treated with guinea-pig’s liver cells. The gastrotoxin has not been removed from the serum by the liver cells, and the stomach therefore shows a patch of necrosis of the mucous membrane. The action of the serum has, however, been weakened. Fie. 4.—The lower stomach is that of a guinea-pig which was injected with gastrotoxic serum previously treated with guinea-pig’s liver cells. The gastrotoxin has been removed in this case by the liver cells, and the stomach therefore shows no lesion. The upper stomach is from the control animal, which was injected with untreated serum, and shows lesions which are relatively slight. PLATE 17. Fic. 5.—Illustrates the fact that guinea-pig’s red blood corpuscles will not remove the gastrotoxin from the serum by combining with it. Bolton. Roy. Soc. Proc., B. vol. 77, Plate 16. [Photographs by B, S, Worrall.] Bolton. Roy. Soc. Proc., B. vol. 77, Plate 17. [Photographs by E. S. Worrall.) 1906.| Specificity and Action in Vitro of Gastrotoain. 44] Upper Stomach.—From a guinea-pig which was injected with gastrotoxic serum previously treated with blood corpuscles. Necrosis of the mucous membrane has resulted, because red blood corpuscles will not combine with the gastrotoxin. Lower Stomach.—From the control animal, which was injected with the same dose of untreated gastrotoxic serum, necrosis of the mucous membrane has therefore resulted. Fic. 6.—-Ilustrates the fact that rabbit’s stomach cells will not remove the gastrotoxin from guinea-pig-rabbit gastrotoxic serum. Upper Stomach.—F rom the control animal, which was injected with untreated serum, and therefore shows extensive necrosis and hemorrhage. Lower Stomach—From a guinea-pig which was injected with gastrotoxic serum, previously treated with rabbit’s stomach cells. The cells have failed * to remove the gastrotoxin from the solution, and lesions similar to those of the control animal have resulted. Fic. 7.—Illustrates the hemolytic lesions which are produced in a guinea-pig’s stomach by the injection of a guinea-pig-rabbit entrotoxic serum. As seen here, the lesions are indistinguishable from those produced by gastretoxic serum. Fic. 8.—Illustrates the fact that guinea-pig’s stomach cells will not remove the hemolysin from guinea-pig-rabbit hemolytic serum (obtained by injecting red blood corpuscles). Hemolytic patches are seen in the stomach, which is that of a guinea-pig. The animal was injected with guinea-pig-rabbit hemolytic serum, prepared by injecting the red blood corpuscles of the guinea-pig into the rabbit. Before its injection the serum was mixed with stomach cells for one hour. The stomach cells have failed to remove the hemolysin from the serum. 442 The Influence of Increased Barometric Pressure on Man.—No. I. By Leonarp HI11, F.R.S., and M. GREENWOOD, Jun., M.R.C.S. Research Scholar of the British Medical Association. (Received January 16,—Read February 15, 1906.) Introduction. The classical researches of Paul Bert,(1) confirmed in recent years by v. Schrotter(2) and his co-workers, and also by Leonard Hill and J.J. R. Macleod (3 and +), have demonstrated beyond question that the ill results observed in caisson workers and divers are to be attributed entirely to injudicious rapidity of decompression. Experiments on animals have shown that every 100 c.c. of blood or tissue fluid dissolve, at body temperature, about 1 c.c. of nitrogen under one atmosphere of air; 2 cc. under two atmospheres; 3 cc. under three atmospheres, and so on (Hill and Macleod, Hill and Ham). (5)* This nitrogen is set free as bubbles in the capillaries and tissue spaces when the decompression period is made too short, and by the embolism of some vessel, may produce symptoms varying in kind and severity. One of us (L. H.) having determined, by numerous experiments on animals, that no ill effects follow exposure to pressures up to +seven atmospheres, if 20 minutes be allowed to each atmosphere for decompression, we determined to investigate the effects of high pressures of air upon ourselves. The records of caisson works and the operations of deep sea divers show that owing to the rapid rates of decompression at present employed by engineers and divers, very great risk is incurred by workers in caissons at pressures of +3 atmospheres, and by divers at depths of from 100 to 150 feet. As, however, divers usually stop a very brief time, while caisson workers outstay a shift of from 2 to 4 hours, the body fluids of the latter become saturated with nitrogen, hence their greater danger at lower pressures. The limit for practical diving work is fixed by the great increase of mortality and illness which occurs at depths much exceeding 100 feet, while at less depths than this, accidents are by no means infrequent; being occasionally very severe or fatal in character. The Admiralty set 120 feet as the limit of work for their divers, while the most daring pearl and sponge fishers sometimes reach depths of 145 feet; in * Bohr (‘ Nagel’s Handb. d. Physiologie,’ 1905, vol. 1, p. 117) gives the coefficient of absorption of arterial blood exposed to an atmosphere of N,, at body temperature as 1°26. The Influence of Increased Barometric Pressure on Man. 443 this latter group accidents are numerous. Lambert, the famous diver employed by Messrs. Siebe and Gorman, salved £100,000 at a depth of about 160 feet. On each descent he passed about 20 minutes below, and about the same time in ascending. On the last journey he stayed longer and became affected on his return to the surface, permanently losing the power to retain his urine. Lambert was the man who stopped the flooding of the Severn Tunnel, going through the tunnel (dark and full of water) in a Fleuss dress to a distance of a quarter of a mile from the shaft, and closing the flood gates, which had been left open; his courage deserves to be recalled. Another diver, Erostabe, salved treasure from a depth of 171 feet, and yet another, Ridyard, from 160 feet. These three divers of Messrs. Siebe and Gorman hold the record for successful work carried out at great depths. Two other divers of the same firm, in order to test a patent kind of diving apparatus, descended to 189 and 192 feet respectively. One of these divers (Walker) tells us he was about 50 minutes over the job, taking 30 minutes to ascend. He ascribes his immunity from accident throughout his career as a deep diver to his habit of slow ascent. The deepest dive on record is one of 204 feet (+884 lbs. pressure); the diver who made this record died from the effects of too rapidly mounting to the surface. In 1894, at Bordeaux, H. Hersent,(7) an engineer in charge of caisson works, having first experimented on animals, found three workmen willing to submit themselves to high pressures of air. These men were enclosed in a steel chamber, and the experiments were conducted under the observation of a commission composed of five members of the Bordeaux Faculty of Medicine. Two of the workmen had had previous experience of compressed air. In one experiment the subject was compressed to +4°800 kilos. per square centimetre (+6827 lbs. per square inch) in 35 minutes, remained under this pressure 1 hour, and was decompressed in 2 hours 3 minutes. On quitting the chamber the man experienced a few “ picotements,” which lasted for half-an-hour, but no other unfavourable symptoms. In a second experiment, a pressure of +5°000 kilos. (+7116 lbs. per square inch) was attained, without any subsequent ill effects beyond a few “ picotements.” Finally, the same subject was compressed to + 5-400 kilos. (+76°81 lbs. per square inch) in 45 minutes, remained under the pressure 1 hour, and was decompressed in 2 hours 25 minutes. The effects are recorded in these words: “A ressenti peu de picotements, cela tient aux bains sulfureux pris les jours précédents.” (8). Hersent’s experiments justify his conclusion that “avec quelques précau- tions en sus de celles qu’on prend ordinairement, les hommes peuvent étre comprimés et décomprimés sans danger pour leur vie, et que méme leur 444 Messrs. L. Hill and M. Greenwood, Jun. [Jan. 16, santé n’est pas menacée quand on atteint des pressions allant jusqu’a 5 kg. 400.” (9) Hersent and his medical colleagues do not appear to have entered the pressure chamber themselves, so that we are not in possession of an accurate record of the subjective effects as noted by trained scientific observers. One of our objects therefore has been to study in detail the subjective and physiological changes mduced by greatly increased barometric pressures ; another object has been the investigation of the respiratory exchange under the same conditions. In the present memoir we shall communicate the results already obtained. Part I. Our experiments have been carried out in a steel cylinder kindly placed at our disposal by Messrs. Siebe and Gorman, the eminent firm of naval engineers, to whom we are further indebted for much valuable assistance. This cylinder (vide photograph, p. 446) had a capacity of 42:2 cubic feet, and was provided with a mattress, blanket, and pillows, enabling the subject to adopt a comfortable attitude. Compression was effected by means of a two-cylinder motor-driven pump, which could raise the pressure to +6 atmospheres in about 40 minutes. Two decompression taps were provided, with fine bores, permitting very careful adjustment of the rate. of escape. The chamber was also fitted with electric light, bell, telephone, and a thick glass observation window; the latter, however, was subsequently covered with a steel shutter for greater security. The pressure was measured by a Bourdon spring gauge, which had been tested for correctness. We shall now give an account of a typical experiment. The description 1s reproduced from notes taken at the time :— Experiment II, 29.11.05. The subject* (M. G.) entered the chamber at 10.40 a.m. In order to avoid any accumulation of CO2, a constant ventilation at the rate of 25 litres per minute was maintained. * The measurements, etc., etc., of the two subjects were: L. H., age 39, weight (in clothes) 873 kilogrammes, height 1°81 metres, vital capacity 3500 c.c., tidal air 510 cc. ; M. G., age 25, weight 53 kilogrammes, height 1°65 metres, vital capacity 4000 c.c., tidal air 300 c.c. Both were in good physical condition. 1906.| Influence of Increased Barometric Pressure on Man. 445 Time. LSE CINEG) Os Pressure. Notes. chamber. 10.40 A.M 57° F. +0 10.50 62 = 10.55 — +16 lbs. Voice becoming nasal and metallic. 11.5 67 = 11.20 69 + 62 lbs. Sensation of slight vertigo. 11.34 68* OZ Ti Between 11.25 and 11.40 articulation was difficult, and the subject experienced some trouble in making himself heard through the telephone. 11.55 — +77 lbs. 12 noon 65 = Subject quite comfortable, voice still nasal but easier to produce and much more audible. 12.4 P.M. = + 72 lbs. j 12.10 — — Pulse, 40. Respirations, 9 per min. 12.37 64 + 52 lbs. 1.0 63 +31 ,, Voice much better. 1.20 634 — Pulse, 42. 1.51 == +0 Period of compression, 54 minutes. Period of decompression, 2 hours 17 minutes. On quitting the chamber some itching was perceived in both forearms,, especially the right. In about 20 minutes neuralgic pains were felt, localised in the radial side of the left forearm. These pains gradually increased in intensity, spreading up the arm; then, after remaining moderately intense for five minutes, they gradually subsided. Several minutes later (about one hour after leaving the chamber) similar pain was experienced in the right forearm. This however did not spread upwards, was less severe and quickly subsided. An hour and a half after leaving the cylinder the subject felt quite well and no subsequent ill effects resulted. As will appear later, there is good reason to suppose that the slight discomfort present at the conclusion of this experiment is attributable to the fact that the subject remained almost completely at rest during decompression. We may therefore conclude that an adult may be safely submitted to a total barometric pressure- of at least 7 atmospheres, which is, we believe, a limit higher than any previously reached. In the course of our investigation the following pressures have been attained :— Subject, L. H. Subject, M. G. 75 lbs., once. 90 lbs., once. 60 ,, twice. 75 ,, three times. AD), Be 602 iour ss 30 ,, four times. AU) Th i 30 ,, seven ,, * Wet cloths were placed on the cylinder at this time. + This reading was verified by Mr. J. A. Craw, who was present during the whole- course of the experiment. [Jan. 16, ll and M. Greenwood, Jun. i Messrs. L. H 446 dvj worssa *OUTAL 3] ata pooLS 5) ‘adid uorssardwi0y ‘yy ‘quoydelay, “aT admooag *q ‘(peso]d) AOPUIA WOTYBAtEsqQ *O {faq OMQooTy “gp ‘HCISNI NVWHUOM HLIM “(NGadO) UMANITAO FO MATA ‘QjoyuRy, “Ww 1906.] Influence of Increased Barometric Pressure on Man. 447 In no case have any severe after effects resulted. The maximum pressure in our series corresponds to a water depth of 210 feet, which is 90 feet beyond the limit fixed by the Admiralty for their divers. Supposing the special diving bell designed by one of us (L. H.) for the slow decompression of divers were employed, it seems quite possible that work might be carried out safely at a depth of 210 feet. Even a greater depth than this might be attained by an intrepid man, for the limit appears to be fixed by the pressure at which the toxic effects of high tension oxygen become an immediate danger. These effects have been studied by Paul Bert, Lorrain Smith (10) and Hill and Macleod (3). When the partial pressure of oxygen reaches 2 atmospheres (corresponding to 10 atmospheres of air, or a depth of about 350 feet of water) convulsions may occur in animals within 20 minutes. The limit of possible safe working is therefore about 250 feet. Conceivably this limit might be extended by diluting the air with nitrogen so as to lower the partial pressure of the oxygen, but we do not claim more than that our experiments show the safe diving depth may be increased up to 210 feet. The responsibility of those who allow short decompression periods in caisson works is clear; every death or case of paralysis from air embolism must be set down to the negligence of the contractor. Next, as to the sensations we felt under pressure: the feeling of discomfort in the ears and deafness, due to a difference in air pressure within and without the tympanum, is too well known to need description. Owing probably to a catarrhal condition, we were unable to open our Eustachian tubes by merely swallowing, and were compelled to resort to a forced expiratory effort with mouth and nose shut, the latter being held tight by the finger and thumb. To one of us (L. H.) who had not practised beforehand the opening of his Eustachian tubes, the first séance was most disturbing. The sensation of increasing deafness and discomfort, more than discomfort, in the ears, with no obvious cause, and the inability to gain relief by the recognised method of swallowing, produced a feelimg of mental distress which led to his signalling to terminate the experiment. Once having learned the method of opening his tubes, no such trouble resulted on subsequent occasions. As to whether one possesses any real sense of the amount of pressure, the answer must be in the negative. V. Schrotter and his co-workers (11), who made observations in caissons sunk in the Danube at from +0°5 to +2°65 atmospheres, say that: “ Bleibt nun der Druck stationir, verweilt die Person auf langere Zeit unter einem bestimmten Drucke, so hort mehr oder minder rasch, off mit einem Schlage, jegliche unangenehme Sensation im VOL, LXXVIIL—B, 2K 448 Messrs. L. Hill and M. Greenwood, Jun. [Jan. 16, Ohre auf, nur das Gefiihl von Dumpfheit, das Gefiihl eines vermehrten Widerstandes im Ohre, wird in der Mehrzahl der Falle, besonders von Ungewohnten, wahrgenommen.” We found that all distinct sensations of pressure in the ears were relieved immediately the pump ceased its strokes, and the pressure in the chamber became constant. Our hearing was as acute and, in the opinion of L, H., more acute than normally. The signal of a tap with an iron spanner on the outside of the chamber was, to L. H., painful in its intensity. Apart from the feelings of nervousness at being exposed to so high a pressure (which at times were somewhat acute, especially when we were not engaged in analytical work), we could not detect any real sense of pressure, and certainly noticed no abnormality in our bodily functions, with the trifling exception of the voice. Thus during Experiment XV _ the subject (L. H.) when at +60 lbs. wrote: “ Very nervous all through experiment; whenever time for thought, the feelings of pressure, if any, due to non-equilibration of ears when pressure is rising.’ During the same experiment, when the subject learnt he was at +55 lbs., he wrote: “Thought one was lower until told. No real sense of pressure except lip and voice change.” In another experiment, M. G. was nearly two atmos- pheres too low in his estimate of the pressure, while in a third experiment made at a period when custom had lessened the nervous effect, he replied to a question at +60 lbs., “ no sense of pressure.” The voice changes, observed in all caisson workers, were well marked in ourselves. The alteration is distinct at +1 atmosphere, and very marked at +3 atmospheres. The voice has a peculiar nasal and metallic quality, losing the individual characteristics of the speaker. Thus to L. H., when speaking in the chamber, under pressure, his voice appeared like that of M. G. under pressure. So close was the resemblance that L. G. could fancy himself outside and listening to M. G. through the telephone. At +3 atmospheres the power to whisper or whistle is almost entirely lost. L.H., who retained the power somewhat longer than M. G., could just make an audible whistling note at this pressure. This loss of the fine vibratile movements of the tongue and lips, a loss probably resulting from the damping effects of the dense air, leads to a false sense of anesthesia in the former parts. This conception of anesthesia is interesting, as being solely excited by a lack of normal movement. V. Schrétter and others have laid stress on the diminished frequency of the pulse and lowered blood pressure of caisson workers. Our observations are not sufficiently extensive to permit of any final pronouncement; but, so faras they go, we are unable to detect any definite change in the pulse 1906.] Influence of Increased Barometric Pressure on Man. 449 frequency. For instance, in Experiment II, M. G.’s pulse was at the rate of 40 per minute at +70 lbs. and 42 at +63 lbs. In Experiment XIV, it was 41 per minute, at +50 lbs., 30 at +30 lbs. 42 at +10 lbs., and 41 at +2°3 lbs. This subject’s pulse is normally slow, being rarely above 60 per minute in the sitting posture; hence although there appears to have been a diminution in frequency, the change is not nearly so striking as in the cases tabulated by V. Schrotter (12).* L. H. found no alteration in his pulse-rate at +5 atmospheres. Our observations on the blood pressure have not been at all complete. The Hill and Barnard pocket sphygmometer, depending as it does upon a column of air acting as an elastic spring, is not a satisfactory instrument for high pressure work, the viscosity of the dense air lessening the excursion of the pulse very greatly. We came to the conclusion that it was an important matter during the decompression to move in turn every muscle and joint of the body, and to change one’s position frequently, so as to keep the capillary circulation active in every part. In the brain, spinal cord, and abdominal organs this circulation is kept active by the work of the respiratory pump. In the limbs, muscles, fat of the back and chest, on the other hand, the movement of the blood and lymph back to the heart depends mostly on changes of posture and the expressive action of contracting muscles. The following observations support these views. 5 In Experiment XIII M. G. was decompressed from +75 lbs. in 95 minutes. During decompression he flexed and extended all the limb joints at frequent intervals, with the exception of the knees. Subsequently pain and stiffness were detected in the knees and nowhere else. In Experiment XIV the same subject was decompressed from +5 atmos- pheres in 120 minutes. During the compression all the limb joints, including the knees, were repeatedly moved. No after effects of any kind were experienced. A further difference between the two experiments was that in the second a pause of about five minutes was made at each atmosphere for analytical purposes. As in each of the experiments followed by pain (in the case of M.G.) no such pauses occurred, it is possible, but we think not probable, that these may also play a part in hindering the development of after effects. The most interesting experiment in this connection is No. XV. L. H. was decompressed from +5 atmospheres in 105 minutes, a pause of five minutes being made at each atmosphere. During the decompression movements of * In Schrotter’s cases there was no direct relation between barometric pressure and pulse frequency. 450 Messrs. L. Hill and M. Greenwood, Jun. [Jan. 16, ) the joints and muscle of the limbs and back were carried out regularly. On emerging from the cylinder, beyond a few “picotements,’ no unpleasant symptoms were noticed. On the next day the subject wrote as follows: “The only place I did not move and massage was the front of the chest, where I have plenty of subcutaneous fat. In the evening painful places were felt in the sub- cutaneous tissues of the anterior thoracic region; one spot under each nipple, one across the right side of the chest about the level of the ensiform cartilage, another above the left axilla in front, and one over the right upper arm in front. A red or purplish rash appeared over these tender places. They felt like a spot in which a subcutaneous injection of water has been made. Next morning the tenderness was better but still evident, and the rash was subsiding.” Forty-eight hours after the experiment this purpuric rash was still discernible, and was shown to Dr. W. Bulloch and other pathologists. An eruption occurred in a very severe case of caisson illness seen by Heller Mager and v. Schrotter (13). They give a plate of the eruption, which is described in these terms: “ Haut der linken Schulter und des linken Armes an der Aussenseite, besonders in der Gegend des Olekranon und des diusseren Condylus sowie in der Gegend des Biceps mit lividen, blaulich- rothen netzformig verzweigten, inselformigen Flecken bedeckt, ebensolche auch am Handriicken.” The arm of this sufferer was much swollen and intensely painful. These observations then show the extreme importance of active movement and massage during decompression ; instructions should be given to all caisson workers to perform such movements while in the air lock. We believe the tenderness and the rash were caused by small bubbles embolising the vessels of the subcutaneous fat in the case of L. H. The pair felt by M. G. was probably due to small bubbles in the nerve sheaths in the first case, in the knee joint in the second. Part II. The next stage of our investigation was devoted to an inquiry as to the changes in the percentage of alveolar CO, under the altered conditions. We have employed the method described by Haldane and Priestley (14) The subject breathes through a wide-bored rubber tube; after a normal expiration he expires deeply and then closes the end of the tube with his tongue. A sample was taken from the wide tube into Haldane’s portable CO, analyser, and examined. A bench fitted up in our cylinder enabled the subject to collect and examine samples with ease. It may be remarked that 1906.| Influence of Increased Barometric Pressure on Man, 451 it is necessary to replace the corks at the bottom of the water bath in Haldane’s apparatus by well-fitting rubber ones, as the air is compressed in the corks, which leak at high pressures. Owing to the loss of the water jacket some of our earlier experiments were unsuccessful. Great care is also necessary in readjusting the potash levelling tube, as when the chamber is closed a slight fall of pressure is almost inevitable owing to escape round the washer of the door. Haldane and Priestley have shown that the respiration is so regulated as to maintain a constant tension of CO, in the alveolar air, which is generally about 5 per cent. of an atmosphere. Now supposing the metabolism to be unaffected by changes of pressure, and the regulation of respiration to continue the same, the amount of CO, in the alveolar air must vary inversely as the pressure attained. Thus if p be the percentage of CO, at normal pressure, then we should have, at two atmospheres, p/3 per cent. of CO. in the alveolar air. It will be seen that these conditions were almost exactly realised by us. The following table gives the result of two typical experiments. The figures in brackets give the percentages reduced to + 0 lb. in accordance with the above principle. Strictly speaking, the exact height of the barometer should have been recorded, and another correction ought to have been made for charges of temperature in the cylinder. As there is however a necessarily large experimental error, we think it needless to allow for these minor differences, and have accordingly assumed the normal atmospheric pressure to be 15 lbs. to the square inch, and neglected the temperature :— Experiment XIV. 10.1.06. Subject, M. G. Percentage of CO in alveolar air. Pressure. 5°3 (5'3) + 0 lbs. 0:9 (5'4) (Mean of two) +75 ,, 1:0 (5:0) +60 ,, 213 (5:0) FoR 1°8 (5:4) +30 ,, 2°7 (54) i BA (54) ae Hxpervment XV. 10.1.06. Subject, L. H. Percentage of CO, in alveolar air. Pressure. 47 (4°7) + 0 lbs. 0°9 (45) . +60 ,, 0-7 and 0°8 (4:5) +75 ,, 0:95 (4:75) +60 ,, 1:2 (4°8) +45 ,, 1°8 (5°4) +30 ,, 2°5 (5:0) +15 ,, 50 (50) eiOuls 452 ~~ Mesers. L. Hill and M. Greenwood, Jun. —‘ [Jan. 16, The next tables comprise all our results. The figures vertically beneath one another refer to the same experiment :— Subject, M. G. Pressure. Alveolar percentages of CO.. lbs. | 53,54 | 5:3 55 5-7 | 536-7 |. 68-be4 8 | 3°3(5°06) | = — | — _ _ -- | | | 2-7 (5-4) — Ot to oo) “- fs cS | | | 16 | | 2-7 (6°58) | 2-7 (6°58) - ws 1°8 (5°4) 1°8 (5°52) | he bo © Fs Pal Tel al exea! eo —— 21-3 (5:2) | 21-3 (5°38) — 1-0 (5:0) & her | Lh We _ @ Wal TS] | oes} OK SF | | 0-9 (4°6) ae 75 | | — | 0°9 (5 “4) Subject, L. H. | Ere Alveolar percentages of CO. | | | l | Ibs. | | | AS) Oo Or & or fo) > a oO r—) ES PL AA eee eA: AS) Oo BS 5 Xt oo boas Ee NH vo a Co on a on Sa =e Ve = (or) or = 258 6 wa rw bo 8 = tl i eS roy oS Hd us pr OU NGS is wae HS a ) i — =~ ie,2) e9) = oo 8 [ees nde ee) Vt Lee oe TORS io Noo @ Ges | toa Wh deta | ————~ . SSS SY I | 15 j We think these results show so close an agreement with the theoretical values as to support the conclusion that changes in the percentage of carbon dioxide in the alveolar air depend solely upon the physical conditions. No increase or decrease in the pulmonary output of CO2 occurs. Metabolism, then, in so far as it can be determined by an investigation of the alveolar air, is not affected by increasing the barometric pressure. It is scarcely neces- sary to add that this criterion is by no means adequate to sustain the jinal conclusion that metabolism is, in fact, unaltered by the atmospheric condi- 1906.] Influence of Increased Barometric Pressure on Man. 458 tions; so far as it goes, however, it is in favour of such an inference. Summing up the results of the present investigation :— It is proved that— (1) A man can be submitted to a total pressure of seven atmospheres without untoward effects, provided decompression be affected gradually, and the capillary circulation be aided by repeated contractions of muscles, joint movements, and charges of posture. (2) We have no sense of increased barometric pressure so long as the former is constant. It is probable— (1) That the subjective effects of increased pressure, apart from voice changes and lip anesthesia, depend upon psychical conditions such as anxiety and excitement, (2) The changes in the percentage of carbon dioxide in the alveolar air are conditioned solely by physical variations, and not by any increase or diminution in the respiratory metabolism. In conclusion we would remark that we are unable to find any evidence in support of Snell’s (15) opinion, that the presence of CO, in the respired air exercises a peculiarly unfavourable influence under increased pressure. Thus in one experiment the percentage of CO. in the chamber air, at + 31 lbs. was 0°62 (equivalent to over 1°8 per cent. at + 0), and no untoward results occurred on decompression. These researches were carried out with the aid of a grant from the Royal Society Government Grant. REFERENCES. 1. Paul Bert. ‘La Pression barométrique.’ Paris, Masson et Cie., 1878. Heller Mager and v. Schrétter. ‘ Luftdruck-Erkrankungen mit besonderer Beriick- sichtigung der sogenannten Caisson-Krankheit.’ Wien, 1900; also v. Schrotter, ‘Der Sauerstoff in der Prophylaxe und Therapie,’ etc. Berlin, Hirschwald, 1904. 3. Hilland Macleod. ‘ Journal of Physiology,’ vol. 29 (1903), No. 6. 4, —— ‘Journal of Hygiene,’ vol. 3, No. 4; ‘Roy. Soc. Proc.,’ vol. 70, p. 455. 5. Hilland Ham. ‘ Physiol. Soc. Proe.,’ July 1, 1905, p. vi. 6 7 Lo Heller Mager and v. Schrétter. Op. cit., pp. 474, et seq. H. Hersent. ‘Note sur ’Emploi del Aircomprimé pour l’Exécution des Ouvrages hydrauliques, et spécialement des Fondations.’ Paris, Imprimerie Chaux, 1895. 8. —— Op. cit., p. 34. Op. cit., p. 21. 10. Lorrain Smith. “ Pathological Effects due to increased Oxygen Tension,” ‘ Journal of Physiology,’ vol. 24, p. 19 (1899). 11. Heller Mager and v, Schrétter. Op. cit., p. 624. Op. cit., p. 658. Op. cit., p. 524 and Plate ITI. 14, Haldane and Priestley. ‘Journal of Physiology,’ May, 1905. 15. Snell. ‘Compressed Air IIness, or so-called Caisson Disease.’ London, 1896, Lewis, p. 212. Studies on Enzyme Action—Inpase. . By Dr. Maurice NicLovux. (Communicated by Professor W. D. Halliburton, F.R.S. Received J ee 16,— Read February 1, 1906.) Ina recent number of the ‘ Proceedings of the Royal Society ’* Dr. Henry G. Armstrong published a paper with the above title. I beg leave to draw attention to the workt I have performed on the saponification of fats by castor-oil seeds, and, without entering into detail, to state my general conclusions. These are as follows :— (a) By mechanical means it is possible to separate the cytoplasm of the eastor-oil seeds from all the other cellular elements, particularly from the aleurone grains. (6) Pure cytoplasm prepared as above alone has the property of hydro- lising fats ; its power is considerable. (c) It acts on the fats in the same way as an enzyme, and follows all the laws of enzyme action. (d) Nevertheless the active substance of which cytoplasm is but probably the support is not an enzyme; this substance, which I proposed to call “ lipaseidine,” is destroyed by water as soon as it is no longer protected by fats. (ec) It is possible to repeat in vitro with isolated cytoplasm hydrolysis of the fatty matter such as occurs in the seed at the time of germination. * “Roy. Soc. Proc.,’ B, vol. 76, p. 606. + These were published in a series of notes in the ‘Comptes Rendus de |’Academie des Sciences’ :—“ Sur un procédé d’isolement des substances cytoplasmiques,” ‘Compt. Rend.,’ 1904, vol. 138, p. 1112 ; “Sur le pouvoir saponifiant de la graine de ricin,” ‘Compt. Rend.,’ 1904, vol. 138, p. 1175; “Etude de Vaction lipolytique du cytoplasma de la graine de ricin,” ‘Compt. Rend.,’ 1904, vol. 138, p. 1288 ; “ La propriété lipolytique du cytoplasma de la graine de ricin n’est pas due 4 un ferment soluble,” ‘Compt. Rend.,’ 1904, vol. 138, p. 1352; “Mécanisme d’action du cytoplasma (lipaseidine) dans la graine en voie de germination ; realisation synthétique de ce mécanisme,” ‘Compt. Rend.,’ 1904, vol. 139, p- 148; and later in a general memoir, “La saponification des corps gras,’ ‘Revue Générale des Sciences,’ 16¢me Année, No. 23, 15 Decembre, 1905, pp. 1029—1037. 4595 On the Function of Silica in the Nutrition of Cereals.—Part I. 3y A. D. Hatt and C. G. T. Morison. (Communicated by Professor H. E. Armstrong, F.R.S. Received December 22, 1905,—Read February 1, 1906.) (From the Lawes Agricultural Trust Committee.) 1. Introduction. The presence of silica in plants was first demonstrated by the analyses of De Saussure,* who pointed out that the Graminee were particularly distinguished by the large proportion of this constituent present in their ash. Liebig, who classified plants as “silica plants,” “lime plants,’ and “potash plants” according to the predominance of one or other of these constituents in their ash, in accordance with his “mineral theory,” regarded the silica as a necessary element in plant nutrition. This view led Wayt+ to introduce as a cereal manure a rocky material derived from the Upper Greensand near Farnham, which contained a considerable proportion of silicate easily soluble in acids. But when Sachs{ succeeded in maturing maize plants in water cultures containing no silica, whereby the proportion of silica in the ash of the mature plant was reduced from the normal 20 per cent. or so to as little as 0-7 per cent., it became evident that silica could no longer be placed in the same category as phosphoric acid and potash as essential elements of plant nutrition, and Jodin§ raised four successive generations of maize in water cultures without any supply of silica beyond that contained in the original seed. Other investigators again showed that the stiffness of cereal straw, which had been attributed to the presence of silica, depends on the develop- ment of the internodes under the influence of such factors as illumination and exposure. Henceforward little or no importance seems to have been attached to the presence of silica, yet, as the following ash analyses show, it forms a constant and considerable proportion in the ashes of certain plants, though it is almost absent from the majority. * “Recherches sur la Végétation,’ Paris, 1804. + ‘Roy. Agric. Sor. Journ.,’ vol. 14, 1853, p. 225. t ‘Flora, 1862, p. 52. § ‘Ann. Agron.,’ vol. 9, 1883, p. 385. VOL. LXXVII.—B. 24 456 Messrs. A. D. Hall and C. G. T. Morison. [Dee. 22, Table I—Percentage of Silica in Ash. per per | cent. | cent: | | Wheat straw (Rothamsted mean)...! 62°1 |} Hops, leaves (Wolff, mean) 21 ‘1 » grain 4 Ola », cones is 17-2 Barley straw , -| 46-0 || Beech leaves i .| &1°0 >» grain a ...| 18°3 |} Larch needles ‘ 22 °5 Oat straw (Wolff, mean)... 46-1 || Calamus Rotang (Wolff, 1 anal.)...) 68-0 | » grain Py ...| 86°3 || Bambusa arundinacea Pe ...| 28°3 Rye grass (Lolium perenne) ,, .| 26°7 || Sphagnum pulustre . .| 61°8 Maize (whole plant) . .| 430 |} Pteris aquilina 3 .| 43-7 Sugar cane a 56:4 || Hquisetum arvense A -| 41 °7 Erica Tetralix “n 48 °4 ( Owing to the inevitable presence of external dust and dirt upon plant material before analysis, it is almost impossible to say whether the small amounts of silica found in the ashes of many other plants are accidental or inherent. But while it has been demonstrated that silica is not essential to the nutrition even of the cereals, it is hardly likely that a material present to the extent of 60 per cent. of the mineral constituents, as in the ash of wheat-straw, can be wholly without use in the economy of the plant. The only experiments, however, which throw any light on its function appear to be those of Wolff and Kreutzhage.* These investigators grew oats in culture solutions of the type usually described as complete, but further divided into three series, receiving soluble silica in considerable quantity, in a small quantity, and not at all. They observed that while the total growth was not much increased by the presence of silica, the pro- portion of grain formed was considerably raised, a precisely similar effect to that brought about by an addition of phosphoric acid to culture solutions deficient in that element. Hence they concluded that the action of silica and of phosphoric acid were in some way related, the former acting, however, indirectly on grain formation by promoting the migration of the food materials. With this exception the possibility that silica plays any part in plant nutrition appears to have been ignored, just as its practical use in the manuring of cereals has been discontinued. Observation, however, of some of the plots at the Rothamsted Experimental Station, which have long been subjected to a manuring with soluble silicates, seemed to show that the question of the function of silica required further consideration, and an * ‘Land. Versuchsstationen,’ vol. 30, 1884, p. 161. 1905.| Function of Silica im the Nutrition of Cereals. - 457 examination of the records indicated at once that the appearances noticed were not accidental, but had persisted from year to year. 2. Field Experiments at Rothamsted with Soluble Silicates. At Rothamsted sodium silicate has been applied as a manure to certain of the experimental plots over long periods of time, and shows regular and well-marked effects. On the permanent grass plots in the Park, which is cut for hay every year, there are two plots receiving similar heavy applications of ammonium salts, phosphates, and potassium, sodium, and magnesium sulphates. One of these, which receives sodium silicate also, yields a crop exceeding by about 10 per cent. the crop of the parallel plot without sodium silicate, taking an average over the last 42 years. It is possible, however, that the weakly-held sodium base has some part in this action, by neutralising the acidity produced in the soil by the continued use of ammonium salts. This difficulty of interpretation does not, however, apply to the barley plots. In Hoos field, on which barley has grown every year since 1852, one series of plots receives sodium nitrate with various combinations of mineral manures, so as to provide plots receiving (1) nitrogen alone; (2) nitrogen and phosphoric acid without potash; (3) nitrogen and potash without phosphoric acid; and (4) a complete manure. Since 1864 one-half of each of these plots has been cross-dressed with sodium silicate; hence the effect of the silicate is seen in conjunction with each of the elements of a complete manure. The average results obtained are set out in Table IT. Table IT. | | Manures per acre. | Average over 41 years, 1864—i904. ia E | | Grai | St Plot. ; rain. raw. Sodium ee Potas- | Sodium | Magne: itrate. | Pnos- aren |sulphate| lees, sue. . . : | m phate. |sulphate.|°"P sulphate. Without} With |Without} With | | silicate. | silicate. | silicate. | silicate. | i | | | para Posse Ib. ewt. | Ib Ib lb. | Bushels.|Bushels.| cwt. | ewt. | er ieaaio => ; — — — 27°3 33°8 | 16°2 Me} |: [a= dey halite c(h ke cine aaa | — 42-2 | 43°5 | 24°6 | 25°8 | angi | ee | a 200 100 | 100 || 286 | 36-4 | 179 | 217 4 275 | 3°o | 200 | 100 100 41 °2 44°53 25°3 27 °6 In this case only a normal amount of nitrogen is supplied in the form of sodium nitrate, a neutral salt, so that there is no acid to be neutralised 2L 2 458 Messrs. A. D. Hall and C. G. T. Morison. __[ Dee. 22, by the soda of the sodium silicate. The beneficial effect of the sodium silicate is chiefly shown on Plots 1 and 3, and there is little gained by its use on Plots 2 and 4. Now Plot 3 is abundantly supplied with alkaline salts in the shape of sodium nitrate and sulphates of sodium, potassium, and magnesium, so the addition of a further supply of sodium in sodium silicate would not be likely to produce any effect. Rather, if the sodium were the active constituent, would its effect be seen on Plot 2, which receives no alkaline salts beyond the sodium in the sodium nitrate common to all the plots. The notable fact is that the effect of the sodium silicate is seen only on the two Plots 1 and 3 suffering from phosphoric acid starvation, because they have been cropped for so many years without the application of any phosphates. The silica, in fact, would seem to partially replace or to do the work -of the superphosphate supplied to Plots 2 and 4. Such an opinion, derived from the yield, may be confirmed by an examina- tion of the plots when approaching ripeness. The most striking feature at that time is the deferred maturity of the barley on the plots without phosphoric acid; they remain of a greener colour, and are still erect at a time when the barley on the normal plots has turned down and begun to yellow for harvest. This ripening effect of phosphoric acid finds a parallel, though to a smaller degree, on the half plots receiving sodium silicate. On Plots 1 and 3, which are without phosphoric acid, the portions receiving sodium silicate are always riper by a few days than the other halves which get neither phosphoric acid nor silica. A series of analyses of the ash of the barley grown on these plots in 1903, a wet and sunless year, and 1904, a normally warm season, also serve to strengthen the idea that the action of the silica is in some way bound up with that of the phosphoric acid in the plant. Table III shows the percentages of phosphoric acid and silica in both grain and straw on the four plots, each of which is subdivided so as to be with and without silica. It will be seen that the lack of phosphoric acid in the manure applied to Plots 1 and 3 is reflected in the diminished proportion of phosphoric acid in the ash of the grain, and still more in the low percentage present in the ash of the straw. When sodium silicate is added to the plots without phosphoric acid the proportion of phosphoric acid in the grain ash rises, but simul- taneously it falls in the straw ash. On the plots receiving phosphoric acid the silicate does not always cause an increase in the percentage of phosphoric acid in the grain ash, though as before it generally diminishes that in the straw ash. On all the plots the sodium silicate causes an increase of silica in the ash 1905.| Function of Silica in the Nutrition of Cereals. 459 of the grain, and particularly in that of the straw, indicating that under the ordinary soil conditions the barley plant does not obtain all the soluble silica it would otherwise appropriate. Table I11—Hoos Field Barley. Nitrogen and Pure Ash per cent. in Dry Matter, and Phosphoric Acid and Silica in Pure Ash. | ! | | | x | Ni il owe Nitrogen, Nitrogen. || pee pas Nene ene | potash, and || Phosphate. potash. || phosphate. [Sea EAN AR SB PAE SSN AO) | | With | | | | | Witt it _ | With _ | With || | With | ay | silica. | Ca ie, || ula | silica. | Only. | silica ee | | | | | || | ! | | Sa ee ie 2s. til Sant eS. 4. | 48. | | } | | Grain. i| | { Nitrogen in dry matter...) 1°63 | 1°57 || 1°50 | 1-50 aye Seay! 1°53 | 1°54 Pure ash TAGs 198 2°27 | 2°37 ESP OG 2°36 | 2°36 Phosphoric acid in pure 35°80 | 37 “74 | 42°27 | 42°64 | 35°54 | 36°11 | 41°83 | 44°31 ash } | Silica in pure ash ......... 14°19 | 18-67 || 16°43 | 20°60 || 15°81 | 18°00 | 16°95 | 19°71 Ratio, P,O; to N ......... | 0-38] 0-48 | 0-64] 0-67 | 0-40] O-44 | 0-64] 0 1904— Nitrogen in dry matter...) 1°79 | 1°72 || 1°52 | 1-46 ||,1°58| 1-73 146) 1°45 Pure ash Fi 1:94} 2°09 | 2°34) 2-41 || 1:97 2°15 || 2°33 2°32 Phosphoric acid in pure 32°19 | 35°29 || 40°16 | 36-40 | 30°96 | 34°16 | 38°82 | 38°46 ash | Silica in pure ash .........! 16°76 | 20-13 19 -62 | 20-75 || 16°45 | 17°47 |, 16°34 | 19°08 Ratio, P.O, toN ......... | 0°35 | 0-43 || 0-62] 0-60 || 0-39! 0-42 | 0-62! 0-62 Straw 1903— | HI Nitrogen in dry TEES 0°53 | 0-43 || 0-43| 0-41 || 0-°56| 0°50 || 0-42| 0-44 Pure ash os 3°66 | 4°80 } 3-41 fe» Gs 86 4:24 | 4°82 | 3:98) 4°73 Phosphoric acid in pure 2°34) 2°40) 4:18; 3°62 2°41 | 2°19 }) 4°02 |° 4°38 ash | | Silica in pure ash ......... | 51-98 | 63-88 | 56°67 | 64-00 || 46-22 | 55-37 || 48-14 | 57°30 Ratio, P20; to N ......... | 0-16 | 0-27 || 0-36] 0-48]! 0-18| 0-22]; 0:38] 0-48 I l 1904— i Nitrogen in dry matter... 0°49 | 0-48 0-40 | 0-42 | 0°50 0-48 \ 0°43 0°45 Pure ash : 4°07 | 5°00} 4:36] 5-09 || 4°61 | 5-29 |) 4:19) 5-01 | Phosphoric acid m pure 2°66 | 2°13 || 4°47) 4:17 2°48 | 2:02 4-78 3°96 ash | | | j Silica in pure ash .........| 44°00 | 52°54 | 47°19 | 51°28 || 35-91 | 44°07 \ 37°43 44°18 Ratio, P.O; toN ........- 0-21 | 0°23 0°49 | O51 |; 0°23 | 0-22 OAT O44 { | { As the application of a soluble silicate lowers the proportion of phosphorie acid in the straw while raising it in the grain, it would seem at first sight to 460 Messrs. A. D. Hall and C. G. T. Morison. [Dee. 22, act by facilitating the migration and utilisation in the grain of the initially small store of phosphoric acid derived from the soil. But such an interpretation of the function of silica is not borne out if the whole amount of phosphoric acid removed by the crop from the soil on each plot be considered, instead of the proportion of phosphoric acid in the ash. It has already been shown that the use of sodium silicate on the no phosphoric acid plots, 1 and 3, results in a considerable increase of crop, and as the grain of this increased crop is somewhat richer and the straw only a. trifle poorer in phosphoric acid than the grain and straw from the non- silicated portions of the plots, it follows that the whole crop manured with silica contains a greater total amount of phosphoric acid derived from the reserves of phosphoric acid in the soil. This extra phosphoric acid derived from the soil is itself sufficient to explain the greater yield brought about by the silicate without attributing to the silica within the plant any specitic action m economising the phosphoric acid there present. If the function of the silica were to replace the phosphoric acid within the plant and enable it to be moved off to the active tissues and used over and over again, the larger crop due to manuring with silicates would not contain any greater amount of phosphoric acid, but the general growth of the plant, ¢.g., the dry matter produced and the nitrogen assimilated, would be increased. Hence the ratio of the phosphoric acid to the dry matter and to the nitrogen would be lowered in proportion to the increased growth, conditions which are not realised in the cases under examination, where indeed the ratio of phosphoric acid to nitrogen is generally slightly raised by the applications of silicate. The results on the other hand indicate that the silicate gives the plant such a stimulus as enables it to develop more vigorously and obtain more phosphoric acid from the soil, and that all the consequences observed follow from the increase of phosphoric acid thus brought about. Wolff and Kreutzhage held that the function of the silica was to enable the plant to make fuller use of whatever phosphoric acid it had obtained from the soil, the Rothamsted results indicate that its action takes place earlier, in stimulating the plant to draw more efficiently upon the vast but dormant reserves of phosphoric acid in the soil. 3. Effect of Silica on the Development of Barley in 1904. In order to study the question more closely it was decided in 1904 to trace the effect of phosphoric acid and silica upon the development of the barley on these plots at regular intervals from the time of flowering onwards. As the effect of phosphoric acid had been most evident in forwarding the maturation of the crop, it was considered that the later period of the growth 1905.| Function of Silica wm the Nutrition of Cereals. 461 of the crop need only be investigated, «c., the period during which the nutrition of the plant from the soil has largely ceased aud assimilation is coming to a standstill, while the materials previously accumulated in the stem and leaf are migrating into the seed. The method adopted was to take a certain number of rows of barley in the middle of each plot and remove the whole plant, as far as possible with the roots intact, for two yards up these rows, at weekly intervals from June 13 until harvest on August 8, or nine times in all. The plants were then air- dried after washing the roots free from soil, the grain when formed was separated from the straw, and both were finally dried in the steam oven, so as to obtain the weight of dry matter ; although dealing with such small areas it, is impossible to make more than a very approximate estimate of the yield per unit area. The dried material was ground, and after determinations of the nitrogen in one portion, the rest was burnt for ash, in which the pure ash, free from sand and charcoal, and the phosphoric acid and silica were determined. The analytical results are set out in Appendix Tables VII to X, from which are derived the various curves of development now to be considered (figs. 1 to 11). Before however proceeding to a consideration of particular cases it will be convenient to trace by means of an average result for all the plots the general course of development in the later stages of the growth of the barley plant. ‘Grams June 13 20 27 July 4 IL 18, 25 Aug. 1 Fic. 1.—Dry Weights. Mean of all plots. Whole Plant and Grain. Fig. 1 shows the mean dry weight of all the eight plots on each date, both of the whole plant and of the grain. The crop attained its maximum weight about July 18—25, after which it remained stationary and probably indeed declined slightly. The result shown for August 1 is clearly exceptional; on 462 Messrs. A. D. Hall and C. G. T. Morison. [Dec. 22, that date several of the plots happened to yield an exceptionally small number of stems on the area harvested. ; For the better calculation of mean results the smoothed curve also shown in fig. 1 was drawn; by combining the smoothed dry weights read off this curve with the true mean percentages at each date were obtained the data contained in Table IV and expressed graphically in fig. 2. Dry matter 2z00grams N 510, IN (grain) Ioograms , R05 { Dry matter CETaIN) gy FO; (grain) Juneig 20 27 ~— duly Il 18 25 Aug Fic. 2.—Dry Matter, Nitrogen, Phosphoric Acid, and Silica in Whole Plant and Grain. Means calculated on smoothed weights of whole Plant. (SiO, on 3 scale of N and P,O,.) From these curves it will be seen that the dry matter of the crop goes on increasing until about a fortnight before cutting, but the whole of the nitrogen would appear to have entered by July 11, a fortnight before there was any sensible grain. The phosphoric acid seems to reach its maximum at a slightly later date, and the figures for the silica, though subject to greater errors of determination, show that the assimilation of silica continues still later, until the grain has progressed somewhat. Of the nitrogen within the plant, about 63 per cent. is eventually moved into the grain and rather less than 70 per cent. of the phosphoric acid: the migration of the phosphoric acid does not, however, take place exactly pavi passu with that of the nitrogen, but follows it somewhat. Of the silica but a small proportion, 9 per cent. at the maximum, reaches the grain, and nearly the whole of this is transferred in the earlier stages of grain formation, being doubtless present in the adherent pales and glumes and not in the seed proper. 463 919.8 Z89- & 669: & 88S. § OSF- 86-0 FI6-0 -T86-0 C88: 0 e88- 0 4 0gs: T F68 898-1 TSZ- T 999-T | ~~ 3 09g-€ | e088 | ZePrs 89%: & 6g: S| S 962-0 | 30-0 | 288.0 Z9F- O 19-0 | oS 019:0 | 601-0 PEs. 0 916: 0 PST | | > 928-0 | 618-0 | 442-0 0LZ- 0 SsI-0 | = 169-0 | 219-0 | ¥65-0 €ZF- 0 883-0 S O9L- T G8I-T | £0-1 LL. 0 BLP: 0 ~ | = #921 | $-83I| 8 ZET 0. LEI 6: 8ST ~ 9-64 | 9b &- 19 0-09 | 1.9% > 00% 80% S08 L61 cst v ‘raqqyeyy Aq, jo Ss WRN AIT F 2 BLS. & S19. % RF G G22: Z ZL0- & = FEZ. 0 cgz- 0 182: 0 FI-0 | L0F-0 S 08s: 0 Bus. £09- 0 799-0 | Sh2-0 ‘= “= SF. 0 8ZF- 0 ZIF- 0 OFS: 0 BSL 0 cs) 998- 0 123-0 €88- 0 CPs. 0 IT6: 0 Sty 91S: T 6&- T geo. T Oss. T 808: T ~ a) S$ : | a | ‘g ‘any | ‘T‘Sny fa hone IT Ang | : 806-2 | 299-2 PPS-0 -&GL-0 PLL-T | 89-1 808: @ 199: & 8EL. 0 &S4.0 oes. T g89.T OOT- 0 Sy 90T- 0 = GFE: O = I- 6ST €ST 6-01 | ae OLT CT T&T OOT Hor os quer eTou AA, snp osigonpaIGACN eefloe [enjoy “ se $J00L PUB MBIYS ee tye uasoT4I [enV eee e eee eeeeeee Penne eee eee eee 2019 “ popat npooa9sdaosoousoonDnGdE HYALAa «6 UBUD ee ta 0 apet akae aE aa uesoaq1U [eujyoy seereroveseoreeurrirrescereers saooa pus ARIAS AT¢ SO nC See eee eee ured AIT |(pagdope raion poyjoours) .to9qvur Aap [v0], SIYSIO A, POYyQOOWY UO pozwTMoTVo szys1o A i} g9b- T SPL. T POP: O G6P- 0 696: 0 TOL: T 616: 0 | sar 916: 0 — | PEG: SG ce | “p Atup | "yz oung €L8: T 609: 0 GLé- T | ‘0g oun | "eT oun GES: T ¢L9- 0 Cro: T “19qqeyy Arq ut (suvout ony) osvquddI0g | eh eI tet prov ontoydsoyg, §4OO.L pus AVIS es eee ree eene oe WOGOAYT NT ees ea teeco acileranaceagete secuccvencceve eveveeene prox oroydsoy gq WIG) emma war oe | | (19490804 uaye} s}o[g Tv) Aepeg jo uorpisodmoy pue yYySIO A Uvapy 1905. | ‘FOG Uosvag “KopreG plely SooH—' AT 14, 464 ' Messrs. A. D. Hall and C. G. T. Morison. [Dec. 22, Perhaps the most important point brought out is that the grain establishes a particular composition at an early stage in its development, after which, although it continues to grow and increase in weight, it does not sensibly alter its composition. From July 18 onwards the percentage of nitrogen and the percentage of phosphoric acid in the grain remain approximately constant, though the grain gains a further 50 per cent. of its weight during the same period. Whatever chemical changes take place during the latter stages of ripening, they consist in the rearrangement of the minerals within the grain rather than in any progressive change in the character of the intake. Grams 300 pie 2 25): es Whole is plant 200 I 100 2 28 : foe i Junera 20 27 July 4. IL 18 25 Augi 8 Fie. 3.—-Dry Weights of plots without Potash. Grams 200 45 Whole plant 200 3 100 a 3° VGrain 4 3 IL June13 20 27 July 4. i 18 25 Aug. 8 Fig. 4.—Dry weights of plots with Potash. Taking these mean figures as indicating the normal course of development it will now be possible to review the results yielded by individual plots and 1905.| Function of Silica in the Nutrition of Cereals. 465 trace the effect of silica on the assimilation of carbon (dry matter yield), nitrogen, and phosphoric acid, and particularly on the movement of these materials into the grain. Figs. 3 and 4 show in graphic form the yield from individual plots, fig. 3 deals only with Plots 1 and 2, where no potash is supplied in the manure, while fig. 4 deals with Plots 3 and 4, each of which receives equal amounts of sulphates of potassium, sodium and magnesium. In each figure curves are drawn separately for the silicated and non-silicated portions of the plot. The accidental fluctuations in yield from week to week are too violent to admit of smoothing, but the general character of the curves shows that Plots 1 and 3, which receive neither phosphoric acid nor silica, give con- sistently a much lower yield than the others. The curves representing Plot 2 (with phosphoric acid), Plot 1 5S (with silica), and Plot 2 8 (with both phosphoric acid and silica), do not differ from one another by more than the extent of the accidental fluctuations from week to week of any one of them, but all indicate a yield about half as large again as that of Plot 1. Similarly where potash is used: Plot 3, without silica or phosphoric acid, never yields much more than half the crop on the Plots 3S, 4, and 458, receiving either silica or phosphoric acid, or both together. As judged then by the dry matter produced, the silicate manuring is able to do the same work for the plant as the phosphatic manuring on Plots 3 and 4. Despite the magnitude of the accidental fluctuations some differences in the character of the curves may be discerned; both Plots 1 and 3 (without silica or phosphoric agid) reach their maximum only on August 8, whereas in five of the other six cases where silica and phosphoric acid form part of the manure the maximum is reached by July 18 or 25. This would confirm the appearance in the field of deferred maturity in the absence of either phosphoric acid or silica. Fig. 5 shows tle proportion the grain bears to the whole plant at weekly intervals for the four plots which receive no potash, together with the smoothed mean of all the plots for comparison. It will at once be seen that on Plot 1, receiving neither phosphoric acid nor silica, the proportion of grain is below the normal, and also that the grain is later in forming. The 3 per cent. or so indicated on July 4, the earliest date when any separation of grain was possible, would be wholly made up of the adherent pales. It is only in the following week that the weight of grain has become sensible on Plot 1. On Plot 2, receiving phosphoric acid, the formation of grain precedes, and also is finally somewhat above the normal. Plot 1S, receiving silica but not phosphoric acid, occupies an intermediate position; though starting a little later than Plot 2, it eventually becomes 466 Messrs. A. D. Hall and C. G. T. Morison. [Dee. 22, almost identical with it. In other words, the free supply of silica without phosphoric acid to Plot 1S has enabled the plant to mature as high a proportion of grain, and almost as rapidly, as does the supply of phosphoric acid to Plot 2. The further addition of silica to phosphoric acid as on Plot 2 S does not effect any change in the character of the development of the grain. 40% 30% 20% July 4 IL 18 25 Aug 8 Fic. 5.—Percentage of Grain in Plant. Plots without Potash. 30% 20% July 4 Fie. 6.—Dry Grain in 100 Total Dry. Plots with Potash. 1905. | ‘unction of Silica in the Nutrition of Cereals. 467 In fig. 6 curves are seen representing the same succession of plots, this time however they all receive potassic manure. Again, the proportion of grain on the plot without either phosphoric acid or silica, 3, is low, and its formation is retarded as compared with the normal. Plots 4 and 458, the two plots receiving phosphoric acid, are practically identical and agree closely with the normal, while the curve representing Plot 38, where silica but no phosphoric acid is used, occupies an intermediate position. The development of grain on these plots receiving potash is later than is normal, though ultimately as high a proportion of grain to straw is produced. As regards the formation of grain, the curves show that phosphoric acid hastens the formation of grain, and eventually causes a higher proportion of the material in the plant to pass over into that state, while silica acts in the same direction, though not to so large an extent. 25% Junei3 20 2q July 4 II 18 25 Avert Fic. 7.—Percentages of Nitrogen in Dry Matter. Turning to the entry of the nitrogen, fig. 7 shows the percentage of nitrogen in the grain and straw at the successive dates and for the four plots receiving no potash, the mean results being also plotted for purposes of comparison. Plot 1, receiving neither phosphoric acid nor silica, yields the highest proportion of nitrogen in both grain and straw at each stage of the 468 Messrs. A. D. Hall and C. G. T. Morison. [Dee. 22, growth. The use of phosphoric acid on Plot 2 reduces the percentage of nitrogen in both grain and straw to a little lower than normal ‘level, and this reduction is most marked in the grain. Again, silica without phosphoric acid on Plot 1S gives rise to an intermediate curve of development, nearer to the normal than to the curve representing the plot without either phosphoric acid or silica. Silica added to phosphoric acid (Plot 2S compared with 2) makes practically no difference in the curve. July 4 Il 18 25 Aug. I 8 Fic. 8.—Nitrogen. Percentage of whole content present in the Grain. Plots without Potash. Fig. 8 shows the movement of the nitrogen into the grain; although both the grain and straw of Plot 1, without phosphoric acid or silica, contain the highest percentages of nitrogen, yet the proportion of the nitrogen within the plant which passes over to the grain is lower on this plot than on the normal; the transfer again begins at a somewhat later date. The phos- phoric acid alone on Plot 2 induces both an earlier and a greater propor- tionate transfer of nitrogen to the grain than the normal. Silica on Plot 1S induces an earlier and more complete transfer of nitrogen, though not to the extent caused by the phosphoric acid. On the corresponding plots with potash (fig. 9) very similar results obtain; without phosphoric acid or silica (Plot 3) the transfer of nitrogen to the grain lags behind the normal, while the use of phosphoric acid (Plot 4) accelerates this process beyond the normal, silica (Plot 3) acts in the same direction though not to the same extent. 1905.| Function of Silica in the Nutrition of Cereals. 469 80% 60% July 4: ria 18 25 Aug 8 Fic. 9.—Nitrogen. Percentage of whole content present in Grain. Plots with Potash. As regards the phosphoric acid, the proportion of phosphoric acid in the dry matter of the grain is increased by the use of phosphatic manure, as it is also by the use of silica, especially where no phosphatic manuring takes. place. The removal of the phosphoric acid to the grain is naturally more complete in the cases of phosphoric acid starvation ; and when silica with- out phosphoric acid has been supplied, almost the whole of the extra 80% Vig. 10.— Phosphoric Acid. Percentage of whole content present in Grain. Plots: without Potash. 470 Messrs. A. D. Hall and C. G. T. Morison. [Dee. 22, phosphoric acid which the plant had thus been able to acquire is moved off into the grain. This may be seen more clearly in fig. 10, which shows what proportion of the plant’s phosphoric acid is to be found in the grain on the successive dates. On Plot 1, without phosphoric acid or silica, the move- ment of phosphoric acid to the grain begins much later, but is ultimately more complete than on the normal or on the plots receiving phosphoric acid. With silica but no phosphoric acid (Plot 1S) the migration of phosphoric acid begins at an earlier date and the proportion transferred is much increased, in spite of the fact that the actual amount of phosphoric acid in the plant is also much greater than on the first plot. Exactly the same conclusions are derived from an examination of the curves yielded by the parallel plots receiving potash (fig. 11); the use of silica both accelerates the migration of phosphoric acid to the grain and makes it more complete, although a greater proportion is initially present. duly 4 IL 18 25 Augi 8 Fig. 11.—Phosphoric Acid. Percentage of whole content present in Grain. Plots with Potash. The fact that a greater proportion of the phosphoric acid present in the plant is utilised in the grain on the silica plots, seems for the first time to indicate some specific action of the silica in facilitating the migration of phosphoric acid, so that it is not left unused as waste material in parts of the plant no longer active. But it will be found that the actual percentage of phosphoric acid finally left in the dry matter of the straw is no lower where silica has been used, on Plots 1S and 35, than on the corresponding Plots 1 and 3 without either phosphoric acid or silica. If something like 1905.] Function of Silica in the Nutrition of Cereals. 471 0:11 per cent. of phosphoric acid be taken as the lower limit of phosphoric acid in the straw, that limit is just as much attained in the absence as.in the presence of silica. The greater share of the plant’s phosphoric acid trans- ferred to the grain in the latter case comes from the fact that the amount of phosphoric acid assimilated, though increased by the silica, is still not sufficient for the requirements of the plant in the formation of grain, hence the straw continues to be depleted of its phosphoric acid to the lowest limit possible, The consideration then of each of the factors submitted to detailed examination—the formation of grain and the migration of nitrogen and phosphoric acid into the grain—leads to the same general conclusion, that an abundant supply of soluble silica renders the barley plant more able to obtain a stock of phosphoric acid from the soil. On the plots therefore which are suffering from phosphoric acid starvation the manuring with sodium silicate acts like a supply of phosphoric acid; indeed, the plant does actually thereby obtain a larger amount of phosphoric acid. Further evidence that the silica acts by stimulating the plant to take up phosphoric acid is derived from water cultures grown in 1904, Three plants of barley were grown in each of four jars, holding about 3 litres of solution, containing the following nutrient salts per litre :— (@alctumnitrates ss cnacds.ncacnsccesccer ss 1:64 gram. Di-hydrogen potassium phosphate... 0°29 ,, Magnesium sulphate (crystallised)... 0°62 _,, Rotassiumechloratenscce.cestcnccssseasens Ovjleme with a trace of ferric chloride. Growth was vigorous from the first; the barley plants tillered freely and made a large number of shoots from each grain. On June 7 the nutrient solution was replaced by distilled water, which was changed again on the 8th, and replaced on the 9th by a fresh solution. The new solution contained calcium nitrate, magnesium sulphate, and potassium chloride as before in all the jars; the phosphoric acid, however, was varied as follows :— No, 1. No phosphoric acid. No. 2. No phosphoric acid, but 0°146 gramme silica in solution. No. 3. 0°355 gramme phosphoric acid, no silica. No, 4. 0°355 gramme phosphoric acid, and 0°146 gramme silica in solution. It soon became evident that the phosphoric acid and silica, both separately and together, had a ripening effect, which was indicated by an earlier and an increased formation of ears, VOL, LXXVII.—B, 2M 472 - Messrs. A. D. Hall and C. G. T. Morison. [Dee. 22, On August 5, although the plants were by no means fully mature, it was necessary to harvest them because of an attack of aphis. When dried they gave the following results :— Table V.—Barley in Water Cultures. Yield on August 5, 1904. Dry matter. Plot Number of Number of ov. 5 ears. grains. Grain. Straw and roots. Total. grammes. grammes. grammes. 1 4 0 0 35°17 85°17 2 7 5 (0) 31°87 81 ‘87 3 18 177 5°26 50°79 56 -05 4 27 272 9 ‘63 56 °26 65 ‘89 Assuming that on June 9, when the treatment was varied, all of the plants were approximately equal, it will be seen that the extra phosphoric acid added to Nos. 3 and 4 allowed them to double their weight during the remaining period of growth. The silica alone added to No. 2 did little to enable the plant to make better use of the restricted amount of phosphoric acid already in the plant, for although the formation of ears seems to have been a little forwarded, the few grains that were produced possessed no sensible weight. When, however, silica is provided in the presence of phos- phoric acid, No. 4 compared with No. 3, it brings about a considerable increase of growth and an accelerated formation of grain—just such a change, in fact, as would be brought about by an increased assimilation of phosphoric acid. In fact, these cultures demonstrate that although silica cannot replace phos- phoric acid, nor even economise and make more effective a restricted supply already within the plant, it will stimulate the plant to assimilate a greater amount of phosphoric acid should: that be obtainable from the medium in which the plant is growing. Hence, when applied to a silica plant on a soil impoverished in phosphoric acid, it has the same effect in increasing and accelerating the formation of fruit as would result from a direct application of phosphoric acid. It might be supposed that the action takes place within the soil itself, that the sodium silicate in some way attacks the insoluble phosphates of the soil so as to render them more available for the plant, much as an application of lime or gypsum will liberate an increased supply of potash from the soil. On chemical grounds it is difficult to see how such an action should occur, nor do the results with water cultures bear out such a view. To obtain further evidence on this point, samples of soil from the eight plots in question were 1905.] Function of Silica in the Nutrition of Cereals. 473 extracted (1) with strong hydrochloric acid and (2) with a 1 per cent. solution of citric acid. While there is no method of determining the real amount of plant food in the soil which is at the service of the crop, the latter method* gives comparative estimates which are of value when dealing with soils of the same type. Table VI shows a series of determinations of the total phosphoric acid and of the phosphoric acid soluble in 1 per cent. citric acid solution. Table VI. Phosphoric acid soluble in Hotel phosphevieiacid. 1 per cent. citric acid. No silica. With silica, No silica. | With silica. 0 -0086 0 :0067 0 -:0495 0:0721 0:0075 0 :0094 0 ‘0674 0 :0743 Comparing the soils with and without silica, the use of silica has not affected the amount of total phosphoric acid; the greater draft it occasions year by year from the soil of Plots 1 and 3, which are not supplied with phosphoric acid, is barely visible as yet in the analyses. The silica has also little or no effect on the phosphoric acid soluble in citric acid on the four Plots 1,158, 3, and 38; but the amount going into solution is distinctly higher on Plots 2S and 48 than on Plots 2 and 4, all plots receiving phosphoric acid in the manure. It is not, however, on these plots, but on Plots 1 and 3 that the silica shows any effect on the crop, hence these determinations support the conclusion that the sodium silicate has no action upon the soil phosphates. Though the seat of the action is thus transferred from the soil to the plant, it is by no means settled whether the stimulus which the silica gives to the plant to enable it to take up more phosphoric acid from the soil reserves is a general stimulus or a specific one confined to the phosphoric acid. In other words, does the presence of a free supply of soluble silica so invigorate the plant that it is enabled to repair any weak link in the chain of nutrition and get as need be more nitrogen, phosphoric acid, or potash from the soil, or is the beneficial effect confined to the phosphoric acid alone? It is chiefly towards the settlement of this point that the further experiments both with silica and non-silica plants are now being directed. * Dyer, ‘Chem. Soe. Trans.,’ vol. 65, 1894, p. 115. 2M 2 A474 Messrs. A. D. Hall and C. G. T. Morison. [Dec. 22, The further question of the intimate mechanism by which the silica acts within the plant, and the nature of the chemical changes into which it enters to bring about the observed effects, cannot yet be raised. In the first place little is known of how the phosphoric acid itself acts; it is evident that it induces seed-formation and hastens maturity, but in what way it takes part in the cell processes is still doubtful. Some of the data accumulated in the present investigation may profitably bear discussion in this connection— it is evident, for example, that there is little or no interdependence between the phosphoric acid and the assimilation or migration of nitrogen, as has often been suggested. Again, the results would seem to indicate that a distinction must be drawn between physiological maturity and ripeness, The grains of a phosphoric acid starved Plot like No, 1 go through a ripening process, but they never approach to the composition, or even attain the appearance, of the truly mature grain on more normal Plots like 2 and 4, The grain from Plots 1 and 3, though ripe, has still many of the characters of immature grain, Ifthe progress of the grain be judged by such factors as the per- centage of nitrogen or the ratio of phosphoric acid to nitrogen, the grain early in its formation settles down to a standard composition correlated with the original supply of nutriment, and after this point has been reached it does not change its gross composition, though it is continually increasing in size and weight. For example, the grain of Plot.1, with its high percentage of nitrogen and low ratio of phosphoric acid to nitrogen, which might be taken as indicative of its generally immature character, shows no tendency as its grows and ripens to approximate in composition to the thoroughly mature grain of Plot 2. The later stages of ripening are without doubt attended by changes in the nature of hoth the carbohydrate and the proteid contents of the grain, which however are not apparent in the elementary analysis of the grain. Conclusions. The following general conclusions have been reached in the course of this investigation :— (1) Silica, though not an essential constituent of plant food, does play a part in the nutrition of cereal plants, like barley, which contain normally a considerable proportion of silica in their ash. (2) The effect of a free supply of soluble silica manifests itself in an increased and earlier formation of grain, and is thus similar to the effect of phosphoric acid. (3) The silica acts by causing an increased assimilation of phosphoric 1905.] Function of Silica inthe Nutrition of Cereals. 475 acid by the plant, to which phosphoric acid the observed effects are due. There is no evidence that the silica within the plant causes a more | thorough utilisation of the phosphoric acid that has already been assimilated, or itself promotes the migration of food materials from the straw to the grain. (4) The seat of the action is within the plant and not in the soil. APPENDIX. Table VII.—Hoos Field Barley, 1904. Actual Dry Weights of Grain and Total Plant. No silica. With silica. Date | a of sample. | Plot1.'; Plot 2. | Plot 3. | Plot 18.| Plot 2S.| Plot 38. Nitrogen | No No ete Nitrogen No No é lot ce AE DO: only. | potash. phosphate. Baal only. potash. |phosphate.| ~° Death } Whole Plant. grammes. | grammes. | grammes. | grammes. || grammes. | grammes.| grammes. grammes.; grammes. | June 13...) 84°5 134,°5 53°1 135 °3 140 -4 101 °5 109-4 | 146°7 | 113-2 «2D bod) TRO) 187 ‘5 70°71 180 ‘6 163 ‘0 186 ‘3 146°9 | 164°9 1514 ed 107 ‘9 218 °5 96 °5 186 °3 184-7 216 4, 195 +1 202 -0 175°9 | July 4 116 °5 236 °5 103 ‘0 201-1 194-0 203 ‘8 193 -7 184 ‘7 179-2 | = ilo) USS) 225-0 1040 244, °5 226-6 280-7 194 ‘9 258 -0 209-1 | oy UG) Te 2151 134-7 266 -9 241 +4, 304-0 212-2 255 ‘6 219 °3 - on eB 161 *4 279 -2 139-3 263 °1 245 °1 219-5 260 -2 260-0 228 -5 Aug. 1 158 9 260 ‘9 111 ‘1 226 -5 223 -0 244, °8 139 °4 214 °3 197 -4 me. 8 173 °3 270-2 154 °4 217-1 +|| 211-4 249 -9 248 -7 301°1 || 228°3 Grain. July 4... 3:2 15°8 12°5 16 *4 7-6 14°6 || 11°5 op dlced| TB 41.°7 29°5 49 *6 18-9 41°4 || 29°5 eelSes| 262 71:0 679 85 ‘0 40 °3 69 °1 55 °6 » 25...) 50°4 87-7 84-0 70°6 80 °5 85:4 || 74:9 Aug. 1...) 53°8 83 °3 88 °3 94,°9 48 -7 (oe 72-4, Fs mpiSs 63 °8 74, °2 80 *4 95-6 92-7 108 -0 83 °9 476 Messrs. A. D. Hall and C. G. T. Morison. [Dec. 22, Table VIII.—Hoos Field Barley. Season 1904. Percentage of Nitrogen in the Dry Matter. No silica. With silica. Date : of sample.| Plot 1. | Plot 2. Plot 3. Plot 18. | Plot 28.| Plot 38. Nitrogen No No ene Nitrogen No No ae lot ae only. potash. | phosphate. oe only. potash. | phosphate. or In Grain. July 4. 2 °245 2133 2 358 2 °228 2225 2-193 2 547 2-165 op able 2-198 I 604 2-009 1 “742 1-963 1-670 2-242 1-743 oy lls 1-735 1-431 1°715 1-516 1 ‘637 1 -460 1-739 1 -480 5 ocala S78) 1-451 1 ‘687 1-420 1 566 1-416 1 ‘677 1 504 Aug. 1. 1-944 1 ‘561 1 560 1-467 1-762 1 -430 1-747 1-424 fasts) 1-791 1 ‘517 1-578 1-461 1°716 1-463 1-733 1-448 In Straw and Roots. June 18...| 2:039 1 °421 2 404 1 627 1 ‘524 1-604 1524 1-579 >», 20...) 1°541 1 309 1-551 1 268 1 398 1-234 1 °465 1°413 py PMlece|| Ib BEKO) 1 ‘031 1-261 1-016 1-133 1:012 1-165 1-063 July 4...| 1°162 0°797 1°151 0 ‘977 0 :957 0-909 17121 0-812 » 1...) 0°943 0 646 0-863 0 ‘658 0 ‘796 0-655 0 *862 0 ‘703 >» 18...| 0°868 0 ‘617 0 °824 0 °535 0-704 0-580 0 “747 0 624 » 25...| 0-709 0-544 0 ‘736 0 542 0 599 0 ‘591 0 ‘661 0 546 Aug. 1...| 0-687 0-498 0 626 0 °476 0 626 0 *44.4, 0 640 0 ‘537 » 8...| 0°633 0-472 0 ‘573 0-528 0 538 0-467 0 °542 0 °533 Table [X.—Hoos Field Barley. Season 1904. Percentage of Phosphoric Acid in Dry Matter. No silica. With silica. Date of sample.| Plot 1. Plot 2. Plot 3. Plot 18. |} Plot 2S.} Plot 3S. Nitrogen No No Ghani Nitrogen No No a a2 Pi only. potash. | phosphate, | ~°™P'€ve- only. potash. | phosphate. a In Grain. , July 4...) 0-591 | 1:064 | 0-746 | 0-869 || 0-861 | 1-092 } 1°116 | 1-007 » ll..| 0-908 | 0-787 | 0:872 || 0-946 | 0-932 | 0-932 | 0-911 » 18...| 0°403 | 0°889 | 0-736 | 0:857 || 0-814 | 0-937 |. 0-889 | 0-893. | 4, 25..., 0°572 | 0-923 | 0-693 | 0-962 || 0-845 | 1:007'| 0-879 | 0-959 Aug. 1... 0°631 | 0°945 | 0-591 | 0898 || 0-669 | 0872 | 0-827 | 0-928 » 8...| 0°650 | 0-972 | 0:638 | 0-990 || 0-776 | 0-961 | 0-820 | 0-942 In Straw and Roots. | June 13...| 0°394 | 0-700 | 0-369 | 0-776 || 0-613 | 0°887 | 0°555 | 0-835 » 20...| 0°362 | 0-667 | 0-386 | 0-713 || 0-576 | 0-692 | 0-490 | 0-739 » 27..| 0°306 | 0-586 | 0-308 | 0-553 || 0-442 | 0°603 | 0-388 | 0°556 July 4... 0:256 | 0-543 | 0:294 | 0-511 || 0-401 | 0-654 | 0:342 | 0°545 » 11...) 0°230 | 0-497 | 0-261 | 0-483 || 0-356 | 0-483 | 0-301 | 0-474 » 18..., 0°180 | 0-404 | 0-173 | 0-364 || 0-193 | 0-416 | 0:249 | 0:392 » 25... 0123 | 0°352 | 0-140 | 0-379 || 0-163 | O-416 | 0-161 | 0-408 Aug. 1..., 0°120 | 0-317 | 0-121 | 0-308 || 0-112 | 0-294 | 0-165 | 0°317 HlBets| 20-105 s| '90)-268)m NNO 12275) 032204100116 | mOL308 0-134 | 0°364 1905.] Function of Silica in the Nutrition of Cereals. 477 Table X.—Hoos Field Barley. Season 1904. Percentage of Silica in Dry Matter. No silica. With silica. Date | of sample.| Plot 1. Plot 2. Plot 3. Plot 18. | Plot 2S.|} Plot 38. : Nitrogen No No re Nitrogen No No za oe cre only. potash. | phosphate. | ~O™P'°*|| only. potash. | phosphate. ae In Grain. July 4.... 0°331 0 920 0 °344 0 :730 0 ‘876 1-216 0 °657 1 +236 a5 aba _— 0 ‘680 1-016 0-620 0-800 0°712 0 ‘696 0°777 5 alae CORRS 0 442 0 ‘501 0 621 0-466 0 ‘631 0 -486 0-630 “5 25...| 0-294 0°329 0 448 0-336 0 -484 0 536 0 °384 0-493 Aug. 1...) 0°301 0-403 0 °382 0 :388 0-433 0-521 0 337 0 558 5 ee | 0 °339 0°475 0-339 0°417 0 443 0 548 0 °419 0-468 In Straw and Roots. June 13...{ 1°541 1 523 1-619 1°516 1-850 2-686 1-937 1-969 » 20...) 1°506 1°515 1 ‘764 1 ‘596 2161 2-487 1°932 1 843 op 2Xloell 1724 1 502 1-299 1-259 1-762 2°391 1-881 2 °084 July 4...) 1°191 1 528 1368 0-907 1-988 2 °784 1-767 2 '257 » Ll...| 1°415 1:974 1 604 1 °578 2 516 2-527 2124 2, :280 » 18...| 1643 1 ‘966 1-404 1 569 2° 656 2-905 2 °388 2 558 » 20... 1°809 2-214 2°085 1-954 2-678 3 069 2-462 3 °291 Aug. 1...) 2°210 2 °203 1-858 1°915 2-985 3°550 2 644 3°171 5) Shoal) aL Sai 1-690 2 °356 2 °033 3-028 3 874 2 °748 2-923 478 On Innervation of Antagonistic Muscles. Ninth Note.—Successive Spinal Induction. By C. 8S. SHERRINGTON, F.R.S. (Received January 31,—Read February 15, 1906.) (Physiology Laboratory, University of Liverpool.) It was previously* pointed out that in various reflex reactions inhibition is succeeded by marked exaltation of activity in the arcs inhibited. This after- effect may be figured as a sort of rebound from inhibition. - An example is the following. When a dog in which the spinal cord ts been transected in the thoracic region is, the period of shock having “passed, supported so that its spine is vertical and its hind limbs hang freely, these latter begin to perform a rhythmic stepping movement. This is the reflex, termed by Goltz the mark-time reflex.. The tempo of this stepping differs, in my experience, in different dogs and at different times in the same dog. It may be as frequent at 22 steps of each leg per 10 seconds or.as slow as seven steps in that period. It will persist in some animals for 20 minutes at a time. After, some minutes’ duration its amplitude usually becomes less and the movement on the whole less regular. For the first minutes of duration it is however regular and shows little variation. The stimulus which excites this reflex has not been traced with exactitude. It persists after severance of the sciatic trunk not including the hamstring nerve. Freusbergt inclined to attribute it to afferents belonging to the “muscular sense,” and especially to those connected with parts put under strain in the passive attitude given to the limb under its own weight. It is closely similar to the stepping reflex studied by Philippsont in the dog sup- ported with spine horizontal. That it is initiated by the stretch of some tissue above the knee and especially on the flexor aspect of the hip may be argued from its immediate cessation when the dependent limb is supported from drooping by lifting the lower end of the thigh slightly from underneath by a prop placed just above the knee. Such support, in my experience, usually causes cessation of the reflex in the unsupported (fig. 1) as well as in the supported limb and it does not matter which of the two limbs is supported. The main stimulus, therefore, seems bilateral in origin, and to lie above the * “Roy. Soc. Proc.,’ B, vol. 76, p. 160. + ‘Pfliiger’s Archiv,’ vol. 8. + Heger's ‘Travaux de Laboratoire,’ Bruxelles. On Innervation of Antagonistic Muscles. 479 knee on the flexor aspect of the hip. The attachment of small weights to the foot has not, in my experience, increased the reflex. In the stepping reflex obtained when the animal is supported vertically (the “ mark-time” reflex) the movement is more pronounced at hip than at knee and ankle. A very similar stepping reflex occurs also when the animal is nearly supine. In this latter the movement is more marked at ankle and knee than at hip. In this posture of the animal passive dorsi-flexion of one ankle often excites dorsi-flexion of the opposite ankle, followed by extension at that knee and then by plantar-flexion at that ankle. Re AOAC \ BAN Kove Nua \ NERS LANAAAAN MOQ AAA AAA AAA WA KAN AAA Time tn secs. | | S—— Fic. 1.—“ Mark-time ” reflex of spinal dog. The up strokes correspond with flexions of the limb, the down strokes with extensions. For the period between the two marks on the signal line the reflex was interrupted by taking the limb’s weight off the fellow limb to that yielding the record, namely by supporting it under the knee. On return of the reflex, when the limb was again allowed to hang under its own weight, the reflex shows no increase beyond its previous activity. The small undulations during the period of rest are due to slight swaying of the animal ; the reflex ceased completely. Time is registered above in seconds. There is also a stepping reflex elicitable from the spinal dog when lying on its side and without any marked mechanical strain either of flexion or exten- sion; this is obtained by faradisation of the skin of the opposite hind foot.* * Sherrington, ‘Journ. of Physiol.,’ vol. 33. 480 Prof. C. 8. Sherrington. [Jan. 31, Here the tempo of the stepping is also about 20 steps per 10 seconds, but the stepping is strictly unilateral. These points argue that several sources of excitation probably co-operate in the production of the stepping reflex. An important item in the execution of the movement of the reflex in all its forms is flexion at the hip and knee. Suppose the “ mark-time” reflex is in regular progress and is being recorded from one knee, ¢g., right by a thread passing thence to a pulley and light lever, if then the other thigh (left) be gently supported from behind the knee’ the record shows that the stepping reflex usually at once ceases in the right limb (fig. 1. The refiex ceases entirely: the small undulations on the trace in the interval during the cessation are due to swaying of the body, partly respiratory, in the suspended attitude). The limb during the cessation of the reflex hangs somewhat extended. On removing the slight support from under the left knee the “mark-time” reflex at once recommences, with flexion in the right knee. The reflex, on recommencing after this pause, continues as it ceased, that is, its tempo and amplitude are practically the same as before the interruption (fig. 1). This result contrasts with the following. Goltzand Freusberg* showed that the “ mark-time” reflex can be cut short by a strong squeeze of the tail. In my experience this stimulus is best applied near the root of the tail. A light touch on the hair of the tail often increases the stepping reflex, and the stronger the mechanical stimulus to the tail the quicker and more powerful as a rule is the inhibition of the stepping. But the stimulus to the tail need not be very strong in order to cause inhibition. I judge that the intensity of the mechanical stimulus which, applied to the tail, inhibits the reflex stepping is such that, were the condition of the animal not spinal, would constitute a dolorous (pathic) stimulus. The tail stimulus which inhibits may, therefore, be considered adequate for a nociceptive reaction.f The application of this stimulus to the tail does not in any way interfere mechanically with the stepping movement. Suppose the “mark-time” reflex to be in regular progress and recorded as before, if then the tail stimulus be applied the stepping reflex is almost immediately arrested, and in both limbs. The reflex remains in abeyance while the tail stimulus is continued. On the cessation of the latter the reflex returns, and on its return soon shows indubitable increase in activity as compared with its activity before the inhibitory arrest (fig. 2). The increase is chiefly seen in the amplitude of the movement, but there is also often marked quickening of the tempo of the rhythm. I have seen the rhythm on some occasions quickened by * ° Pfliiger’s Archiv,’ vol. 8. + ‘Journ. of. Physiol.,’ vol. 30, p. 39, 1903. 1906. | On Innervation of Antagonistic Muscles. 481 30 per cent. The after-increase of the reflex may persist in evidence for many seconds. Its decline is gradual. UUURURURURERECURUCEUCUURRUCERERECECERUURUURURUUUURRRUUUURUCEURUUGE: Time wn sees Sugral 'Fia. 2.—“ Mark-time” reflex as before ; but the reflex is here interrupted by stimulation of the tail. This arrest, due to enhibition, is followed, after cessation of the inhibitory stimulus, by increase in amplitude and / slightly in frequence of the reflex. The signal registers period of application of inhibitory stimulus. registered in seconds. ; The arrest of the stepping reflex by tail inhibition cannot be prolonged indefinitely. The reflex tends to return in spite of the inhibitory stimulation when the latter is long persisted in. It is different when the stepping reflex is arrested by lifting one knee; the reflex does not then tend to break through the arrest, however long the latter be continued. In this form of arrest of the reflex the arrest seems referable simply to cessation of the stimulus which excites the reflex. In the case of arrest by tail inhibition the arrest seems referable to a central inhibition, the peripheral stimulus, E, Time 482 Prof. C. S. Sherrington. [Jan. 31, excitatory of the reflex remaining in action allthe time, though unable to produce the reflex owing to the intervening inhibition. The after-increase which ensues, in the second form of arrest, but not in the first, might be explicable in either of two ways. It might be due to the continuance of the exciting stimulus, E, during the period of arrest. That stimulus might, though unable to evoke discharge of the motor neurone during the inhibition, yet be charging a relay apparatus in the reflex are, and so lead to increased discharge after the inhibition was past. Or the after-increase might proceed as a direct result from the inhibition itself, the depressed activity of inhibition being followed by a rebound to super- activity, and altogether apart from the continuance of any excitatory stimulus during the inhibitory period. To decide between these possibilities the effect of strongly stimulating the tail when at the same time both hind limbs were supported from below was tried. The stimulus for the stepping reflex was thus held in abeyance at the time of and during the whole period of the intercurrent inhibition. The result was found to be an after-increase of the stepping reflex not less marked than in the previous cases. It is not at first obvious what relation a stimulus to the tail bears to the reflex of the limb. But it is often noticeable that in the “mark-time” reflex the tail itself is alternately deflected to right and left, keeping time with the stepping reflex. When the right limb begins to draw up in flexion, and the left limb to straighten out in extension, the tail begins to move from the right to the left. The tail does really therefore participate in the locomotor reflex, of which the stepping movement is also a part. Nocuous stimuli to the side of the tail, eg., by unipolar faradisation, evoke reflex abduction of the tail from the side stimulated, and the organ is then usually kept abducted for a time, just as the hind paw is drawn up and kept so for a time when excited by similar stimulation. Moreover, such stimulation of the tail excites reflex movement not only of the tail but of the hind limb, and the limb’s movement is usually extension at hip and knee. The result of thisis, that the tail stimulus can inhibit a flexion-reflex of the hind limb. If the flexion-reflex be induced by inserting a hedgehog spine into the planta, and if while that prolonged reflex is in progress and the limb is remaining thoroughly flexed at hip and knee, a caudal skin-point is faradised, the imb at once drops into the extended attitude under gravity. This occurs when the nerves to the extensors to the hip and knee have been severed. The tail stimulus therefore inhibits the flexors of knee and hip. The particular mode in which the tail-stimulus comes to inhibit the stepping reflex seems to be that it inhibits the rhythmic flexion of the hip, 1906. | On Innervation of Antagonistic Muscles. 483 which is so prominent a part of the mark-time reflex. The after-increase of the latter reflex following on its inhibition by the tail-stimulus seems, as shown above, a pure effect of rebound from inhibition. On the above view it should show itself therefore chiefly in an after-increase of the flexion of hip movement, and the graphic records of the effect show this to be the case (fig. 2), the movement of flexion being the upward movement in the tracing. This inhibition of the “mark-time” reflex exemplifies therefore the principle of the common path,* The reflex arc whose reaction is inhibited and the reflex are which inhibits are both found when separately examined to use the same final common path, but to different effect. The common path in this case is the flexor neurone of the hip, and one arc uses it in a steady depressor manner and the other in a rhythmic pressor manner, The conflict in this case, as so often, is between a nociceptive reaction and a purely locomotor reaction; and the former prevails as is usual,f The after-increase consequent upon inhibition is evidently a form of “bahnung.”t In order to distinguish it from those forms of “ bahnuny” which ensue without previous inhibition and are therefore immediate, it may be conveniently termed “ successive spinal induction,’ the more so as that term draws attention to the likeness between the spinal process and certain visual phenomena commonly designated “ induction,” Another instance of “successive spinal induction” is the following: In the spinal animal (cat, dog) lying supine, the knee-jerk is elicited at regular interval by tapping the patellar tendon. If, then, the central end of the previously severed hamstring nerve is faradised, the knee-jerks become much less ample or quite inelicitable. The tonus of the knee-jerk muscle (vasto- erureus) is at the same time depressed. On discontinuing the stimulation of the hamstring nerve, the knee-jerk quickly becomes again elicitable, and soon is more brisk and ample than prior to the intercurrent inhibition§ (tig. 3). The tonus also returns and in some cases becomes clearly greater than prior to the inhibition. This after-increase of the knee-jerk takes place when, during the whole period of inhibition, the leg is by mechanical support prevented from drooping, and thus the passive stretch of the vasto-crureus is avoided during the inhibition. The after-increase also occurs when the elicitation of the knee-jerk is completely remitted during the whole period of the inhibition. The after-increase is not therefore due to any continuance * Sherrington, ‘ Brit. Assoc. Reports,’ 1904, Address to Section I, + Ibid. ¢ Exner, ‘ Pfliiger’s Archiv,’ vol. 28. § ‘Roy. Soc. Proc.,’ B, vol. 76, p. 161. 484 Prof. C. 8. Sherrington. [Jan. 31, of the action of exciting stimuli during the period of inhibition. It is a “successive spinal induction” following upon inhibition, just as in the instance previously given. \ BESeeeeanau \_\ eaaual \\ WEEEEEEEEEEREUY AA AL VASA AAA h JS \ \. MA wl Time in secs Fig. 3.—Knee-jerks. The knee-jerks were elicited by taps of equal intensity delivered at intervals signalled by a metronome. During the time marked by the signal the afferent nerve of a flexor muscle of the knee was weakly faradised. This inhibitory stimulus depressed the knee-jerk. After the inhibitory stimulus was discontinued the jerks increased to beyond their amplitude prior to the inhibition ; this increase is accompanied by a tonic after-action following each jerk. Time registered above in seconds. In the “ scratch-reflex,” after its inhibition by the crossed extension-reflex, or the homonymous flexion-reflex, a similar after-exaltation is sometimes seen. Fig. 4 exemplifies such an occurrence. But the time of interruption of the reflex has usually in my records been too short to allow much scope for the development of successive spinal induction, and the quick tiring of the scratch-reflex under electric excitation is unfavourable to examining it there. 1906. | On Innervation of Antagonistic Muscles. 485 aN Time tn-2” Stgnal B Stgnal A Fic. 4.—Scratch-reflex of spinal dog. The reflex was evoked from the skin of the shoulder by unipolar faradisation applied for a period marked by the signal line A. The refiex soon after its commencement was inhibited by stimulating (unipolar faradisation) the skin of the planta of the foot engaged in the scratching-reflex. The time of the inhibitory stimulus is registered by signal line B. After cessation of the inhibitory stimulus the scratch-reflex—its stimulus being continued throughout—returned, and on return was more ample than before the inhibition. Time registered above in fifths of second. 486 Prof. C. S. Sherrington. [Jan, 31, It is easy to evoke reflex-extension of the hind limb by stimulation of the skin of the opposite hind limb. With the spinal dog laid on its side (e.g., left) and a thread attaching the knee of the slightly flexed right limb to a recording lever, the delivery of a certain stimulus (eg., 15 break-shock at 30 per second) by unipolar faradisation through a gilt needle at a skin-point of the left foot, evokes reflex-extension at right hip and knee, If this stimulus, at moderate and unchanged intensity, be given at regular interval (e.g., once per minute) a series of extension reflexes of regular height and duration are obtained. If in the course of such a series the right limb is, during one of the 60-second intervals, thrown into strong reflex flexion (¢g., by faradisation of the skin of its own foot and the reflex flexion be maintained for some time, ¢.g., 40 seconds) the next extension-reflex following on the intercurrent flexion differs from those prior to it in being more ample and more prolonged (figs. 5 and 6), Its after-discharge is greatly increased and its latency is sometimes diminished. If the test stimulus for the extension-reflex be adjusted at just subliminal value, the intercurrent flexion-reflex will make it supraliminal. The exaltation of the extension-reflex may remain perceptible for five minutes; in the example furnished in fig. 5 it is quite recognisable for four minutes. The mode of production of this exaltation seems the same as that traced above in the knee-extensor, with the knee-jerk as test-sign, and in the stepping reflex after interruption by tail inhibition. It has been shown that central inhibition of the extensor of the knee is part and parcel of the homonymous flexion-reflex of the leg. In the present case, therefore, during the intercurrent flexion-reflex, the reflex arc of extension was under inhibitory depression. After discontinuance of that inhibition the extensor reflex is found exalted to a degree of activity beyond that which it showed prior to the inhibition. A similar successive spinal induction is evident in the following. The spinal animal (cat) being supine, the nerves of one hind limb are severed save for the nerve to vasto-crureus. The limb is supported with femur vertical and anticrus therefore fairly flexed at knee. The central end of the hamstring nerve is faradised; this causes the usual reflex inhibition of vasto-crureus, an inhibition which, if there be little or no tonus in that muscle at the time, is difficult to detect by mere inspection, though easily revealed by abolition of the knee-jerk, If the faradisation of the central end of the hamstring nerve be continued for some seconds, eg., three, on its cessation there often ensues a marked reflex extension of the knee. This is no mere return to previously existent slight tonus; it is a fairly intense contraction of the vasto-crureus, often sufficient to extend the knee fully 1906. | On Innervation of Antagonistic Muscles. 487 and passing off again in three or four seconds’ time, It sets in, in my experience, not at the very moment of withdrawal of the inhibitory stimulus, but in the course usually of the first three seconds following that with- drawal. A tap given to the patellar tendon appears. sometimes to elicit it when otherwise it would not ensue spontaneously. Related to this phenomenon seems the following. When a flexion reflex of the hind limb is by appropriate stimulation continued for a long time in the spinal dog, the flexion tends to be broken through™* from time to time by short-lasting explosive extensions of the limb, much resembling the “extensor-thrust.” In all these cases the extensor arc during the fiexion- reflex has been under prolonged inhibition, and the superactivity which it shows under a test stimulus, the “ spontaneous” discharge which it exhibits on relief from the inhibition, and the explosive outbreak which it gives when the inhibitory reflex is getting fatigued, all seem to be evidences of “successive spinal induction” supervening as a rebound after inhibition. The effect upon the direct flexion-reflex of an intercurrent extension reflex is, in my experience, much less marked than the converse just described. This may be due in part to my having used the crossed extension-reflex and not a direct extension-reflex as the intercurrent reflex. The crossed reflex is less potent and powerful than the uncrossed reflex.t But the only homonymous extension-reflex of the limb available is the “extensor-thrust,” and that is so unmanageable, and especially is so little capable of prolongation, that it was unsuited to this purpose. However, an influence can be traced, and in other ways than by intercurrent extension in a series of reflex flexions. Thus, with the spinal animal vertical, the hind limbs are taken and kept fully extended at hip, knee, and ankle; then usually, in a short time, a strong flexion-reflex at hip and knee supervenes. Again, if similarly one hind limb be strongly passively flexed at hip but strongly extended at knee and kept in that posture for a short time, it is usual for any attempt to passively extend the hip to elicit at once strong reflex contraction of the flexors of the hip, preventing passive extension. Nevertheless, the greater inductive effect of flexion upon extension than of extension upon flexion as examined at the knee-joint, seems, in my experience, marked. In regard to it one remembers that though electrical stimulation of the afferent nerve-fibres from the flexor muscles has been shown to inhibit the reflex contraction of the extensor,t it has not been * ‘Journ. of Physiol.,’ vol. 34, p. 34, fig. 21, phase 3. + Sherrington, ‘ Brit. Assoc. Reports,’ 2bzd. t ‘Roy. Soc. Proc.,’ vol. 52, p. 556. VOL. LXXVIIL—B. : 2N ‘Sp1009I JO SOLIOS OY} noYysno.m[y perayjeun pourewesr uorvordde sz Jo qutod aq} pur sninuys oy} Jo AzIsuaqUT ay, ‘spUooes UT MOTEq poT]|VUSIs st OWL, ‘snTNUITYs Yyoue Surui10j syooys-yveIq Jo solsos oy S10qSIFe1 OAOG” [VUSIS OY, *XOYoI-Uorxey oy} Jo UoZE[VorSzUT OY} eA0Joq poyzrorye ‘o[duexe ue sl _V YOIyA Jo ‘sottes oy} ul Surureyqo 4eyy 07 [enbo qnoqe st Aq1sueqUI xoyer oy} pur “Guesord you sl MoIyeyUNOMUsNe ayy ‘xoYoI-UOIxey poye[eo1oqUl oY} 109ze OINUIM YFG oy} UI payoao ‘| Xopor UT ‘S[VAIEZU eynUIUI-aU0 4e poezorfe os[e “| pue “q ‘H Soxoyor Suraop[oz ceaqy ey} Ut snoraqo si ynq ‘AT[enpeas soepisqns esvotoul sy, ‘asaeyostp-10jye ut Ayjeroedse ‘aoryeanp puv opunyifduie ut osvetoUl SA\OYs XOYoI-UOTXey pozeTwo1eqUl oY} 104Je 4xou SULMOT[OJ q Xoper UoIsuezxe oy, ‘SpMooes cg JO paurTezureUI pue peyoaoid sem ‘xoPYoI-UOISM9}xe Pesso.1d ay1 Ul WOIsta}xe 07UL SultozUe quiT, yey} Jo ATotueu ‘quIT[ ey} Jo Xopfol-UOlxey SuOT4S Bg PUL VW SUOTZL[NUITYS UseAJoq “PUodes-JTeY BURY} S89] PoqSe| ‘S]UALIN oY} Jo WOIsse0dns oY} oJ UasoY aye ay} 4e ‘sN[NUITS 94} Yeqy Os ‘paxeAaTfap Suteq OT ATUO pur “Yrom Suteq syueTIM9 peonput ay} ‘AzIsueqUI MOT Jo SEM Xoyer oY, ‘pxooar oy} Surprard yeqy 07 GUMMY MOT[ey eq} Jo 4ISIP v Jo ULyS oY} 07 ATreTodruN podde §}UA1INd-poonpur yeeriq Jo setres yenbe Aq sTearezUl oyNUIW-9U0 4V poxOAe SVM Xoo oy, “Sop jeurds oy} Ul Xoyot-WoIsu9}xe PEssol]—'¢ “OI ington H = o -S oD) S. . C of. er 1906. | On Innervation of Antagonistic Muscles. Wl Fie. 6.—Crossed extension-reflex. The reflex was being elicited regularly by short series of break-shocks, the series being equal in intensity and duration. (In 26 nine shocks were delivered instead of eight in the other stimulations, owing to a defect in the rotating key.) Reflex No. 1 is the last of a series of equal reflexes thus provoked ; both stimulus and reflex were of low intensity. After reflex No. 1 in the one-minute interval between it and 2a, a strong flexion-reflex of the limb was excited and main- tained for 40 seconds. The next following reflex 2a exhibits augmentation, and this augmentation is obvious also in reflex 2b. Reflex 2d was elicited two minutes after 2b, and shows no augmentation. It is somewhat less than the reflex evoked just prior to the intercalated flexion-reflex. The signal above registers the time, etc., of the break-shocks used as stimuli. Time is marked below in seconds. The intensity of the stimulus and the place of its application remained unaltered throughout the series of observations. JEN 490 | Prof. C. 8. Sherrington. [Jan. 31, shown conversely that similar stimulation of the afferent fibres of the extensor muscle (vasto-crureus) inhibits contraction of the flexor muscle. To examine this latter point is not altogether easy, since the nerve severance of the vasto-crureus nerve, in order to stimulate its central end, of necessity renders impossible the maintenance, let alone the examination, of any reflex status of that muscle. I have, however, succeeded in splitting the nerve, and if that is done without too much damage to either half of it, one division can be severed and dissected back into the psoas, and the other remain preserving the knee-jerk and, more important, a fair extensor tonus at the knee. The central end of the severed division can then be stimulated, and its effect on the rest of the extensor muscle itself be thus observed. To employ electrical stimulation with as little risk of escape of current as possible it is necessary to cut the divided nerve as low as possible, and dissect it up into psoas as high as possible; in this way I have obtained sometimes 5 em. length of free nerve. I have further exsected the internal saphenous nerve right up into the psoas, and also the external division of the muscular division of the anterior crural nerve, hoping thus to minimise current escape from the stimulated branch of vasto-crureus nerve to those other branches of the anterior crural. In a previous communication I discarded effects of electrical stimnulation of the nerve of the extensor muscle as too open to error by escape of current.* Further examination with the above precautions makes me, however, regard the following as reliable reflex effects obtainable by stimulation of the central end of the vasto-crureus nerve itself: inhibition of the tonus of vasto-crureus itself, the tonus returning to some extent immediately on cessation of the stimulus if the stimulus be weak and brief; inhibition of the knee-jerk, contraction of rectus femoris, especially of its upper part, and tensor vagine femoris and psoas, dorso-flexion of ankle, some slight contraction of the hamstring muscles, especially deep inner hamstring, and extension of opposite knee with inhibition of its hamstring muscles. These effects obtained by faradic excitation are also obtainable, but in slighter measure, by drawing a ligature tight upon the central end of the nerve (mechanical stimulation). The total effect is flexion of the homonymous and extension of the opposite limb. There is thus no evidence that the afferent nerve-fibres from this extensor muscle when excited in these ways inhibit contraction of the flexors, although the afferent fibres from the knee-flexor when similarly excited do inhibit contraction of this extensor. Reciprocal innervation is evident in the reflex effect obtained from the afferents of each muscle, for those of * “Roy. Soe. Proc.,’ B, vol. 76, p. 283. 1906. | On Innervation of Antagonistic Muscles. 491 each inhibit one set of muscles and excite the muscles antagonistic to the inhibited group. But in each case the reciprocal innervation has the same direction, namely, excitation of the flexors and inhibition of the extensors. This relation would obviously tend to make it more facile for flexion of the limb to successively induce extension than for extension to induce flexion. And another consideration has to be borne in mind. The measure to which the intercurrent flexion-reflex exalts the following crossed extension- reflex can be estimated in terms of the relation existing in the crossed extension-reflex between intensity of exciting stimulus and intensity of reflex response. A somewhat widely-expressed opinion is found in the literature dealing with reflex action to the effect that intensity of reflex response is relatively little determined by increase of intensity of exciting stimulus (Wundt,* Hallstén,t Biedermann,t Baglioni§). Observations by Merzbacher|| in the intact limb of the frog, and by Parif! in the isolated gastrocnemius, by myself** in the scratch-reflex, and by Langendorffff in the flexion-reflex of the tortoise, show that in some spinal reflexes at least there can be obtained marked grading of intensity of reflex response in conformity with grading of intensity of stimulus. In the extension-reflex of the hind limb as obtained by stimulation of the opposite hind foot the amplitude of the movement and its duration increase with increase of intensity of the exciting stimulus. The relation between the intensity of the stimulus and that of the response in this reflex is, in my experience, somewhat different from that which obtains in several other reflexes, e¢.g., in the direct flexion-reflex and in the scratch-reflex. The successive increments of intensity of stimulus cause increase of the extension-reflex by fairly gradual and regular degrees up to a certain point. Beyond that point relatively larger increments of reflex response result from increase in intensity of stimulation (fig. 7). This character of the ratio in this reflex between increment of stimulus and increase of response is especially evident with the after-discharge of the latter. In regard to the successive spinal induction exemplified by the reflex, it is clear that if the intensity of stimulus chosen for testing the crossed extension-reflex be near below that value at which * ‘Untersuch. z. Mechanik d. Nerven u. Nervencentren.’ ‘ Archiv f. Physiol.,’ 1885—1888. t ‘Pfliiger’s Archiv,’ vol. 80. § ‘ Verworn’s Zeitschrift.’ || ‘ Pfliiger’s Archiv.’ * 4 ‘ Archives Italiennes de Biol.’ ** © Physiol. Soc. Proc.,’ March, 1904. tt ‘Sitzungsb. d. Naturforsch. Versam.,’ 1905. + ‘g ‘SY SuIpperd 4eyy se [LUILUe OUTS JY} WIOTJ SI px0d01 OY, *gpU000S Ul MOTEG SUIT, “suTORs oy Ut petaysiSar sp10dat tT} JO vate Jo SPIUN UL XeYer Jo aAnsvaut { a[eos MINTIOJONPUL AaxDoUOTY, OY} JO SPUN UL TaATS ST SNTNTUTYS Jo AZISU9zUT OY, — A C.F . 69h a GcE ay 6 L G C.g ‘Xoffar JO ainsvay] UOTZL[NIUYS yora wsemjaq posdvya soynutut om} Jo [eAsoqUI UW q ro) ais) an Ja el ei o) eI op) oD) ns iS) a jal 492 09¢ OOFL OOTL 068 069 09¢ GLY rar) G “T Xopoed ‘snqnmiys Jo Aste} UT ‘4 SUIAOT[OF JXOU 9UO 9y} pur ‘molqgoosuvay [Vutds ay} Jo paeapvey pordde pue osnyip sem epour oy} £4003 oy1s0ddo 9y} Jo JISIp &% JO WIYS oY4 07 opoyyry o[peeu v Aq pesJoarfep o10M ‘puodes Jed CZ Jo 04" oYy 4v ‘SYoYS-Yeorq ZI Koper owe Jo Wor7VepOIS ey} a0,q ‘snzeredde Suryejnwys oyQ Joy moro Aavwnad oyy ut ‘Teusis yausem-o1joa]e oyy Aq our, doz yy UI UMOYS o7¥ sn[NUII4s 044 jo WOleInp pu swTy sy, ‘sn_NaIIys Jo AzIsuaquT Jo Stipras YIM Surpuodsess00 xepyot AjISUS}UI JO SUIPVIX) “XOYOI-UOIsU9}xXe PoessoIg—"), “D1, 1906. | On Innervation of Antagonistic Muscles. 493 further increment produces great increase of response, the exaltation induced by an intercurrent flexion-reflex need not be very extreme to give, neverthe- less, a very apparent and great increase in the response. A not very extreme superactivity induced in the internal condition of the are might suffice to give the external stimulus a value equivalent to a stimulus that would produce a very much greater reflex response. In the homonymous flexion-reflex the increments of reflex response ensuing from increments of intensity of stimulus follow, in my experience, a more regular progressive increase (fig. 8) than in the crossed extension- reflex. There is therefore with this flexion-reflex less chance of successive spinal induction effecting an augmentation apparently so large as with the extension-reflex. This also has to be remembered, therefore, in contrasting the smaller effect observed in the induction of flexion by extension than of extension by flexion in the hind limb. The linking together of the simpler reflexes which compose a usual reflex eycle doubtless involves several processes; it has attracted the attention of observers from several points of view. Loeb* has illustrated how in regard to segmental reflexes the effect of the reflex in one segment may be to transfer the external stimulus to another segment where it in turn excites the reflex of that segment, and so on further. In this way the reflex sequences, which he terms “ Ketten-reflexe,” can be compounded. Another interesting connecting process welding simpler reflexes into more composite is that discovered by v. Uexkiill} He has shown that a piece of musculature, under static conditions which make it of greater length, is more prone to excitation through the nervous arcs than it is under conditions in which its length is less. Thus, if we suppose a pair of muscles, A and B, which under equal activity retain the lever on which they antagonistically operate in such a position that A is equal in length to B, and if we suppose that a new position be given to the lever such that A is longer than B, the neuro-muscular condition becomes altered so that A is more prone to be excited through the nervous arcs than is B. If I represent rightly in this way the principle arrived at by v. Uexkiill, it will be seen that in some of the experiments mentioned in this Note and in my previous ones, the con- ditions resemble those in which v. Uexkiill finds his principle at work. A third process, qualified to play a part in linking together simpler reflexes so as to form from them reflex cycles of action, seems successive spinal imduction. It appears especially fitted to combine the successive * * Vergleichende Gehirnphysiologie, Leipzig, 1899, p. 96. + ‘Zeitschrift f. Biologie,’ vol. 44. } ‘Sy papperd YOU [BULLY oUUS BYy WOT ‘sSpUOdIS UI MOTEq BUILT, ‘poder of} Jo WOIyBsUepuOD pus uosteduros 10y A Sutovsy eursizo uo pesodaedus puv ‘1equiour YIZy ey SBA A YOIYA Jo Soltos aATJNOasUOd aly} UL PoUIB}qo SpLOoAT [eNyoR oYy WorF sduLOwAy -Koueredsuedy axe SaAdnd 19440 ayy ‘pxooer engow oy UoAT ATJoeAtp poonpoAded oxy St YOITM ‘A GAIN UL Wes [TAM SI OBIBIPOSIP-109ze O[} JO SNUOTO oY, FE 69 * TA | 9 SPL e9LT ! IA | 006 #96 GLEL A xXopou | 1g ¥6 OSes Xopoy *OGABYOSIP-19}JR *xoped “sn NUYS ‘OG AIVYOSIP-19} VV *xoyot *sn{UUtys jo a.msvoyt jo ainsvay jo Aqisuequy | Jo qaunsvoy JO oinseoyy jo Ayisuaquy ‘S[VAIOJUL OJNUIVA-OMY 4B U94ATIM LOpIO OYJ UI PouTEygoO a10M TTA PUB ‘TA ‘A ‘AT ‘TIT ‘TI ‘[ soxopor oy], ‘SeAINO 9s80Y} 07 TOYBIEI UI sN_NUAYS JO UOTZesse0 Jo put AroAT[ap Jo yUoUIOUT 94} AMOYs soatno ydvisoduI O44 UO BssIOsqy ‘snyeredde Suyxynmys oy4 Jo ymoaro Axeuad oy} ut yousem-oxjooqte ue Aq eAoge potoqsider st snTNWTys sy, “puooes tod OF JO 098A 944 4 SyUIAINO-“OTJONpUT yRariq Z) SBA XoYat Yous OF snNUYs oY, ‘sNMUITs Jo AqIsuezUr Jo Surpetd YIM Koper Jo AjIsuezuUL Jo Surperp “sop jeuidgy ~“xKeper-uoIxeT|—'g “Oly —r- 5 aaa SS aa Sherrington. M Oo ei © al 1906. | On Innervation of Antagonistic Muscles. 495 opposite phases of such cyclic reflexes as I have termed “ alternating,’* and shown to be particularly characteristic of the locomotor activity of the mammalian spinal cord. If a reflex, A, not only temporarily inhibits the action of an antagonistic reflex, B, but also as an immediately subsequent result induces in are of B a phase of superactivity, the central organ is in that way predisposed for a second reflex opposite to A to occur in immediate succession to A itself. Such an effect seems proved by the observations in this and a preceding communication. A difficulty in applying it to the case of an ordinary alternating reflex, ¢.g., the stepping reflex of the spinal dog, lies in the intensity and long duration of the reactions which I have employed in order to produce it experimentally. Such intensity and duration certainly do not occur in the course of the alternating reflexes as ordinarily observed. This, I think, does not exclude the likelihood that successive spinal induction is a factor which does contribute to the mechanism of alternating reflexes, although operating in smaller degree than as exemplified in the intenser examples obtained under experimental conditions and mentioned in this Note. Addendum, March 14, 1906. Since concluding the above I have met with marked successive induction and rebound contraction following stimulation of the proximal end of one-half of the split vasto-crureus nerve when the stimulus has been quite brief and weak, «.¢., has not been detectible to the tongue-tip, and has lasted only from 1 to 2 seconds. Starting with the knee in semi-flexion, the stimulus has caused immediate relaxation of the vasto-crureus (inhibition), followed, on > “3 6.—6 7.—4 516 Dr. J. E. Lane-Claypon and Prof. E. H. Starling. [Feb.'12, Interval of 10 days, owing to failure of supply of pregnant animals. March 17.— 7 foetuses, etc., at 16th day. r o1.— 5 A 26th ,, 5 23.— 8 si 22nd _ ,, , 24.—14 i 14th ,, 3 25,—12 - 20th ,, % 27.— 5 20th ,, es 28.— 8 a 22nd ,, : 29.— 5 s Qist ,, 45 30.— 7 55 U7 5, = 31.— 4 ss 20th ,, April 1— 9 5 13th ,, 33 sueng Ee 18th ,, : 5 » 15th ,, 9 6.—11 3 16th ,, is p= B cs 18th ,, This rabbit, therefore, received the fluid extracts of the viscera of 160 foetuses. It was killed on April 8. On reflecting the skin of the abdomen, the mammary glands were seen to be markedly hypertrophied. The margins were almost contiguous, and were somewhat raised and pink, presenting, therefore, much the appearance which is seen in a pregnant rabbit of the eighth or ninth day. The general aspect of the stained gland is shown in fig. 4. On microscopic section not only was there marked duct proliferation with mitotic figures, but at the thickened border the formation of alveoli was just commencing. The whole of these injections had been made intra-peritoneally, or, on one or two occasions, under the skin of the legs, so that there was no infiltration of the connective tissue surrounding the mammary glands. This is the best result which we obtained. Experiment 11. May 2 to June 1, 1905.—In this experiment we sought to determine whether the growth-producing substance is contained chiefly in the viscera or in the body, a.e., muscles, skin, and bones, of the foetus, and, moreover, whether it could be extracted from these tissues by boiling. Unfortunately, however, out of the four rabbits which we chose for this experiment, only two were definitely virgin, so that the results in the other two cases were equivocal. Rabbit 1, not a virgin, received the pressed juice of the viscera. Rabbit 2, a virgin, received the filtered boiled extract of viscera. Rabbit 3, also a virgin, received the pressed juice of the bodies of the foetuses unboiled. Rabbit 4, which was evidently multiparous, received the boiled extract of the bodies of foetuses. From this rabbit, at the commencement of the experiment, a small portion of mammary gland was taken as a control. All four rabbits received portions of 182 foetuses of all ages between May 2 and May 31. From Rabbits 1 and 4 milky fluid could be expressed from the nipples after the ninth injection. In Rabbit 2 a watery fluid could be expressed from the nipples after the seventeenth injection. ; Rabbit 3 showed traces of watery secretion after the twelfth injection. All four were killed on June 1. Results were as follows :— Rabbit 1.—Multiparous. Mammary glands well developed and showing many alveoli on microscopic section. The ducts were full of milk; they were, however, lined with -only a single Jayer of epithelium, and it was impossible to say that any hypertrophy had taken place. Rabbit 2.—Virgin. The mammary glands did not present much enlargement as judged from inspection. On microscopic examination, however, many branching ducts were 1906.| Growth and Activity of the Mammary Glands. 517 observed lined with two layers of cells, presenting the same appearance, but in a smaller degree, as those in the glands of the rabbit in Experiment 10. Rabbit 3.—Virgin. Glands large, hypertrophied, containing a fair amount of watery fluid. Alveoli present and ducts showing proliferation. Rabbit 4.—Not virgin. Mammary glands fully marked, and distended with milky fluid, but impossible to determine whether or not hypertrophied. In order to be certain of the induction of growth in the mammary gland by the injection of extracts of foetus, three more experiments were made. In the first of these, in which the rabbit received 16 injections of the pressed juice of the viscera of 138 foetuses, the results were absolutely negative. In this experiment, however, we had been obtaining very small amounts of pressed juice from the tissues, and we thought that the absence of result might possibly be due either to retention of the active substance by the Kieselguhr or to insufficient destruction of the cells in the process of grinding. It is possible, too, that immaturity of the rabbit may have been in some measure responsible for the negative result. In the next two experiments, therefore, we abandoned the Biichner method and, after grinding with sand and with normal salt solution, centrifuged and filtered the supernatant liquid through a Berkefeld candle before injection. Both these experiments gave positive results. Experiment 12.—October 4 to 21. Virgin rabbit, full-grown. Received daily, intra- peritoneally, the saline extract of the viscera of a number of foetuses about the fifteenth to twentieth day of pregnancy. Killed on the 21st. It showed distinct growth of the mammary glands with duct proliferation (vide.fig. 5). Experiment 13.—October 4 to 21. Virgin rabbit. Received the saline extract, intra- peritoneally, of the bodies and placentze of the same foetuses used in Experiment 12. Fifteen injections were given in the 17 days. Killed on the 21st. It showed marked erowth of mammary glands with plentiful mitotic figures. The appearance of this gland in the stained specimen is shown in fig. 6. Discussion of Results. From the results just described, it will be seen that in six cases we succeeded in producing in virgin rabbits a growth of mammary glands similar to that occurring during the early stages of pregnancy, and consisting in the proliferation of the epithelium lining the ducts, with the multiplication of these ducts by branching into the surrounding tissues. In one of these (Experiment 10) where our injections were carried out during five weeks and the experiment lasted nearly seven weeks, there was an actual formation towards the periphery of the gland of secreting acini. In some of these cases, however, namely those in which the injections had been given under the skin of the back (¢g., Experiment 9), the mammary glands were bathed for considerable periods of time in the injection, and it seemed to us possible that this might be a determining factor in producing growth. We therefore carried out a control experiment on a virgin rabbit, in which normal rabbit’s serum was injected, for the most part subcutaneously, for a period of three weeks. The serum, which was derived from non-pregnant animals, but contained much more nutrient material, ¢c.g., proteid, than the 518 Dr. J. E. Lane-Claypon and Prof. E. H. Starling. [Feb. 12, fluids used in our previous injections, ran down in the subcutaneous tissue, so that during the whole duration of the experiment the abdominal wall was thickened and cedematous through the presence of the serum. On killing the animal at the end of three weeks the glands were little, if any, larger than those usually obtained from a virgin animal (fig. 7). On section, how- ever, mitoses were present in the epithelium of the ducts, and there was apparently a certain amount of proliferation of the ducts. We must conclude, therefore, that superabundant supply of nutrient material in the fluid surrounding the acini may lead to proliferation resembling in kind that which was produced by our injections. ‘This result had not been produced in an earlier control experiment, in which we injected the saline extract of liver, and in view of the small results produced by the injection of the serum as compared with those produced by the injection of the extracts of fetus much poorer in proteids, we are inclined to believe that it is impossible to explain our results in the other experiments as due to the infiltration of the tissues round the glands. This explanation, at any rate, could not hold for the growth in Experiment 10, in which there had been at no time any injection into the subcutaneous tissue of the back. It is interesting, from the general pathological point of view, to note that typical epithelial proliferation in the ducts can be produced by an abnormally large supply of proteid in their surrounding lymph, and the subject is worthy of further investigation. A striking fact in all our experiments with a positive result is the small- ness of the growth produced as compared with the quantity of material used for injection. In all the positive cases the material for injection was derived from foetuses ; in Experiments 11 (2) and 12 from the viscera only ; in Experiment 11 (3) from the bodies only; in Experiments 9 and 13 from the foetuses together with placentz ; and in Experiment 10 from the fcetuses, placenta, and mucous membrane of uterus together. On the other hand. injection of extracts made from ovaries, uterus, or placenta alone had no effect on the growth of the gland. We are therefore justified in concluding that under normal circumstances the hormone which is responsible for the growth of the mammary gland during pregnancy is produced mainly in the growing embryo. This hormone, however, must be produced in minimal quantities. It is apparently not stored up in any of the tissues of the fcetus or of the placenta, so that, in injecting extracts of foetus, we are simply injecting the small amount of material which is diffused through the juices on its way to the blood-vessels and into the maternal blood. It is possible, of course, that the specific mammary hormone is produced from a precursor or mother-substance in some organ or other, and that future 1906.] Growth and Activity of the Mammary Glands. 519 research may reveal some method of splitting off the hormone in large -quantities, and also of determining whether its production is diffused through- -out all the tissues or is confined to one special organ of the body. Injection -of extract of duodenal mucous membrane, for example, would give only minimal effects on the pancreas. We should not be justified in concluding from this absence of result that the duodenum was not the seat of origin of the chemical stimulus to the pancreas. Its peculiar relation to the pancreas is only brought into prominence when it is treated with acid, so as to liberate the secretin from its mother substance. Our experiments, therefore, throw no light on the seat of production of the hormone in the feetus. Apparently the extent of the growth obtained is -a function of the quantity of tissue used in preparing the extracts. The wide- spread occurrence of the substance in the body of the foetus points to its ‘being extremely diffusible, as indeed we should expect from analogy with -other hormones. We can only say, therefore, that the hormone is produced by some or all ‘the tissues of the fertilised ovum, whence it is carried off by the blood to the placenta, and so makes its way by diffusion into the maternal blood-vessels. ‘Whether it is identical with the substances which are responsible for the production of the other changes associated with pregnancy, or whether there are distinct substances acting on each organ which is modified during this condition, our experiments do not show. But we have evidence that in the foetus itself the hormone or hormones of pregnancy have the same result as rin the maternal organism. Thus there is increased growth of the mammary glands in the fcetus during the last month of pregnancy, and also in the ‘female an increase in the uterine mucous membrane, as has been shown by ‘Halban. After birth the mammary glands may begin to secrete just as after pregnancy, and there are changes in the uterine mucous membrane similar to ‘those associated with menstruation. Are we to regard, then, the fcetus as the only source of this hormone ? The facts mentioned at the beginning of this paper show that such a -conclusion is impossible. The growth of the mammary glands which occurs -at puberty can only be ascribed to ovarian influence, and is absent if the ovaries have been previously removed, and Halban ascribes to this ovarian substance both the growth of the mucous membrane during each pro-cestrus -and the swelling of the glands at each cestral period, which may in rare cases be attended or followed by the actual formation of milk. Halban explains in ‘the same way those cases recorded by Heape and Kehrer, in which bitches, which had not been impregnated at the normal time, have, after two months, mot only made a bed for their young, but have had swelling of the mammary VOL. LXXVII.—B. 2 520 Dr. J. E. Lane-Claypon and Prof. E. H. Starling. [Feb. 12, glands, with, in some cases, actual secretion of milk. He would regard this. condition as being a continuance of the state of pro-cstrus leading to: continued growth of mucous membrane and also of the mammary gland. When the impregnation was no longer possible, with the discharge of the: ovum, the secretion of this substance ceased, and the absence of the inhibitory stimulus caused break-down of the uterine mucous membrane as well .as. dissimilative activity of the mammary gland. During sexual life, therefore, the ovaries are continually producing a substance which exerts an influence on both glands and uterus. With the occurrence of conception there is at once a great growth of what we may call germinal material. With the growth of the fertilised ovum the amount of hormone produced in the ovum must also increase in proportion. In the early stages of pregnancy the chief source of this hormone may perhaps be located in the chorionic villi, but with the growth of the body of the fcetus this latter must take a preponderating share in the preparation of the hormone. We have no reason to suppose that the foetal elements of the placenta entirely lose this function of the germinal cells, but the negative results of injection of placente in our experiments show that it is impossible to ascribe to the placenta, as is done by Halban, a preponderating part in the preparation of this hormone. If the hormone is produced in the body of the foetus, it might be objected that the formation should go on after birth, and therefore lead in the new- born animal to a continuance of the growth both of. the mammary gland and of uterus. The profound changes in the environment of the new animal which oceur at birth must, however, induce equally profound changes in its metabolism, and there is no difficulty in imagining that with the assumption of extra-uterine life the formation of this substance in the foetus comes to an end.* The occurrence of growth in the mammary gland of a virgin rabbit and of secretion in the mammary gland of a multiparous rabbit from the injection of boiled extracts of foetus, seems to indicate that the specific hormone, like adrenalin or secretin, is not destroyed by boiling. Further evidence, * In the ornithorhynchus pregnancy is associated with the growth of mammary glands, although the embryo in this animal is contained in an egg, and does not enter into any. anatomical connection with the uterine wall. Halban points out, however, that the shell of the egg is porous, and that during its stay in the uterine cavity it increases in size. and the contained embryo grows, in consequence of the absorption of nutrient material’ from the fluid contained in the uterine cavity. If the embryo is able to absorb nutrient material from the uterine contents, it is equally able to give up to these contents diffusible substances, which may be taken up by the mucous membrane and carried by the circula- tion to the mammary glands. The condition in the ornithorhynchus cannot therefore be regarded as a disproof of the chemical theory which we have adopted throughout this investigation. 1906.] Growth and Activity of the Mammary Glands. 521 however, is required on this point, as also on the question whether the substance is specific to the animal, or whether the injection of extracts of the foetus of one animal would produce a growth of the mammary glands in another species. One experiment, in which we fed a kitten for three weeks on the foetuses of rabbits, was negative in its results. This might, however, have been due to the failure of the intestine to absorb the hormone without destruction, or to the failure of the immature glands to react to the minute stimulus which they received. So far as we know, secretin is not absorbed into the circulation when introduced into the stomach or intestine, and colossal doses of adrenalin have to be given by the mouth in order to produce any systemic effects. The effect of the injections of fcetal extracts on multiparous rabbits deserves some further mention. The multiparous rabbit differs from a virgin rabbit in possessing ready-formed alveoli, 7.v., secretory structures. On the theory which we have adopted, the circulation of the mammary hormone should diminish any secretion in these alveoli and should cause growth. In all our experiments at least 24 hours elapsed between each two injections. It is probable that the hormone was rapidly absorbed from the injection, and was therefore present in the blood of the animal only for a certain fraction, say a few hours, out of the 24. While it was circulating it should cause building up of the secreting cells. Directly, however, it ceased to circulate, the cells would enter into dissimilative activity resulting in secretion. By our injections, therefore, we are not able to imitate the continuous stimulus of pregnancy. We are rather producing each day a pregnancy of a few hours followed by a parturition. These factors should therefore result in the production of milk in any animals possessing the structures (i.c., the alveoli), which are capable of secreting milk, and would therefore account for the secretion of milk observed by us in all the cases where multiparous rabbits were the object of our experiment. CONCLUSIONS. So far as our experiments go, they show that the growth of the mammary glands during pregnancy is due to the action of a specific chemical stimulus produced in the fertilised ovum. The amount of this substance increases with the growth of the foetus, and is therefore largest during the latter half of pregnancy. Lactation is due to the removal of this substance, which must therefore be regarded as exerting an inhibitory infiuence on the gland cells, hindering their secretory activity and furthering their growth. It is probable that the specific substance is diffusible, and will withstand the boiling temperature. §22 Growth and Activity of the Mammary Glands, We cannot, however, claim that these conclusions of ours are firmly established. A final decision can only be given by a research carried on under more favourable conditions. One requires, in fact, a farm, where we could have at our disposal 500 rabbits, and could arrange for a plentiful supply each day of rabbits about the middle of pregnancy. Under these conditions it might be possible to determine both the seat and nature of the effective stimulus, as well as to test the influence of various reagents in splitting off the hormone from some possible precursor. Many of our experiments, carried out in a London laboratory, were brought to a premature conclusion by failure of material. If, however, the conception of the action of the mammary hormone, which was put forward by Hildebrandt and adopted by us, is correct, namely, that it is a substance which produces growth by inhibiting the normal activity of the gland cell, it should be possible to decide many questions affecting it by working on an animal, such as the goat, in lactation. Injection of the hormone should diminish or stop the secretion of milk while it was circulating in the blood, but should, as a secondary effect, produce an increased secretion as a reaction from the immediate assimilatory effect. The injection might, indeed, have to be prolonged for one or two days, since we know that in Man the onset of a renewed pregnancy during lactation stops the flow of milk only after some time (three or four weeks). At any rate, such experiments could be more rapidly carried out than those which have been the subject of this com- munication. DESCRIPTION. OF PLATE. The drawings were made as follows:—The mammary glands were dissected out, pinned on corked rings, hardened in corrosive sublimate and formol, washed, and stained in very dilute hematoxylin. They were then dehydrated, cleared, and mounted as lantern slides in canada balsam between glass plates. (These specimens were shown by projection at the meeting of the Royal Society, on March 1, 1906.) An image of the specimens was thrown (without magnification) on to a piece of millboard, and the darkly stained glands were traced out in indian ink. The figures, therefore, reproduce the glands in natural size. Fie. 1.—Gland from virgin rabbit. 2.—Mammary gland from primiparous rabbit, five days after impregnation. 3.—Mammary gland from primiparous rabbit, nine days after impregnation. 4.—Mammary gland from virgin rabbit which had received injections of extracts of foetuses, uterus, and placente during five weeks (Exp. 10). 5.—Mammary gland of virgin rabbit, showing growth produced by injection of extracts of foetal viscera during a period of 17 days. 6.—Mammary gland of virgin rabbit, showing growth produced by injection of extracts of foetal bodies and placentze over 17 days. 7.—Mammary gland of virgin rabbit, showing slight growth induced by daily subcutaneous injection of rabbit’s serum (from non-pregnant rabbits) during a period of three weeks. 2? aun Lane-Claypon and Starling. Roy. Soc. Proc., B. vol. 77, Plate Fic. 6. Fila. 7. 523 The Internal Anatomy of Stomoxys. By F. Tuttocu, Lieut. R.A.M. Corps. | (Communicated by Professor E. Ray Lankester, F.R.S. Received February 2,— Read March 1, 1906.) The dissections of the local variety of Stomoxys, which form the subject of this Note, were made at the suggestion of Professor Minchin, during his direction of the Royal Society’s Commission on Sleeping Sickness in Entebbe, Uganda. The main object was to furnish some comparison between the internal anatomy of Stomoxys and that of Glossina, and the following notes are based on Professor Minchin’s description of Glossina palpalis. Complete digestion of the human trypanosome seems to occur in 48 hours, in the alimentary canal of Stomoxys; but Lieutenant Gray, R.A.M.C., has found a limited percentage of these Stomoxys to be infected with a Herpetomonas. I am much indebted to Professor Minchin for advice and assistance at every turn, without which these notes could not have been completed. Digestive System. The cesophagus emerges from the chitinous pharynx (which, with the mouth parts, has been described by Hansen) as a flattened tube, which gradually narrows and becomes cylindrical, running at first upwards and then backwards to reach the brain. The connectives of the brain are more vertical than in Glossina. On emerging from their constriction the cesophagus dilates gradually, and runs down to enter the ventral aspect of the proventriculus, which lies in the anterior third of the thorax. The proventriculus (fig. 1, P.) is a mushroom-shaped viscus with a thickened border, and lies with its convexity pointing upwards and slightly forwards. Except for the inversion of its lateral edges, which gives to the proventriculus of Glossina a characteristic outline, the corresponding structure in Stomoxys is very similar in every way. The cesophagus enters the proventriculus a little in front of the centre of its concave ventral surface, and the duct of the sucking stomach running up from below appears to enter with it, though in reality it enters separately at a point immediately behind. As in Glossina, the cesophagus and the duct of the sucking stomach are in the same line. The thoracic intestine (fig. 1, T.I.) arises from the convex dorsal surface of the proventriculus at a point posterior to the entrance of the cesophagus on VOL. LXXVII.—B, 2Q 524 Lieut. F. Tulloch, [Feb. 2, the ventral surface. From its origin the intestine runs down into the abdomen of the fly as a narrow tube of uniform diameter, until it reaches nearly to the lower border of the sucking stomach. At this point it dilates to several times its former diameter, its wall at the same time becoming thinner. The abdominal intestine is proportionately shorter, less coiled, and more distensible than in Glossina; it is about three times as long as the fly itself. The dilated portion of intestine has three simple coils which lie superposed in the middle part of the abdomen, and then gradually narrows, continuing as a uniformly narrow tube downto therectum, The narrow lower intestine has variable bends in its course, but is not coiled. The rectum (figs. 1 and 2, R.) is a dilated cone-shaped portion of intestine, the apex of the cone being towards the anus. Its walls are transparent, and through them are readily seen four long trumpet-shaped papillee, so-called rectal glands, the narrow ends of which are inserted towards the anus (fig. 2, R.P.). A single trachea enters the base of each “gland.” Below the apex of the dilated cone the rectum is continued to the anus as a short narrow tube. In the female this terminal portion of intestine runs within the ovipositor, the anus being situated between the last segment of the ovipositor and the terminal plate. In the male the ejaculatory duct passes over it dorsally from left to right, and runs anteriorly to enter the penis. The appendages of the alimentary canal are the Malpighian tubes, the sucking stomach, and the salivary glands. The Malpighian tubes (fig. 2, M.T., MT.) arise from the narrow lower intestine. The proctodeeum, between their origin and the anus, comprises in length about one-fifth of the abdominal intestine. At their point of origin (figs. 1 and 2, O.) the intestine has a shallow linear constriction. Two tubules arise on each side from a short common tube, and all four tubules are approximately of the same length. The two tubules arising from one side have thickened terminations (fig. 2, T.T.), some four times greater than a salivary gland, and these thickened endings lie in the pericardial sinus. The tubules of the other side are of the same thickness throughout, and their ends lie amid the fat-body of the lower abdomen. Microscopically the tubules are of the usual type. The sucking stomach (fig. 1, S.S.) is a thin-walled sac, made up of one layer of flattened cells with occasional strands of unstriped muscle. It ends at the waist in a very fine duct (D.S.S.) which runs up ventrally to the thoracic intestine and enters the proventriculus (P.) immediately behind the opening of the cesophagus. The alimentary canal and the ducts in the thorax lie in contact with each other in a narrow space between the lateral masses of thoracic muscles. j Cr bo or 1906. ] The Internal Anatomy of Stomocxys. Fic. 1.—Alimentary Canal of Stomoxys. Dorsal view. The thoracic muscles were removed, and the structures in the thorax separated and spread out, though their relative positions are otherwise maintained. In the abdomen the position of the coils of intestine has been very little disturbed, but the Malpighian tubes have been removed by severing their common ducts on each side near O. P., proventriculus; T.I., thoracic intestine; R., rectum; D.S.G., duct of salivary gland ; 8.G., salivary gland; S.S., sucking stomach ; D.S.S., its duct ; O., point of origin of Malpighian tubes; H., the dorsal blood vessel cut short at the heart. The salivary glands (figs. 1 and 2, S.G.) are partly thoracic and partly abdominal. They are comparatively shorter and thicker than in Glossina. In the abdomen they are ventral to the sucking stomach, and from a dorsal view only a knuckle of gland is exposed at the lower border of this viscus. 2Q2 526 Lieut. F. Tulloch. » [Feb. 2, The slightly bulbous ends of the glands lie under the upper border of the sucking stomach, and are found by following up the outer limb of the exposed angle of gland. Except for this angular bend the glands are straight in their whole course, and even when pulled out they are not long enough to reach the hinder end of the fly. . The salivary glands run up through the waist of the fly on either side of the duct of the sucking stomach and ventral to the intestine, and continue with the same thickness to the front of the thorax. At this pomt, in the neck of the fly, the glandular portion ceases abruptly, to be continued as a very fine narrow duct (fig. 1, D.S.G.). At first this duct is made up of small flattened cells, but it almost immediately acquires the structure of a small trachea, becoming chitinised and having similar annular thickenings. At the base of the brain the two ducts join and continue as a single duct on the ventral surface of the chitinous pharynx, inside the transparent membrane which wrapsit round. The dilatationin the common duct which Hansen has. described and which he regards as a storage chamber for the secretion, occurs. about half-way in the length of the common duct. The point of entrance of the duct into the proboscis has been described by Hansen. Nervous System, This consists of the brain and the thoracic ganglion, with the nerves arising from them. The nervous system was not dissected in great detail, but the following nerves were traced as described. The thick nerve to the ocelli arises from the upper part of the back of the brain. The stout nerves to the antenne arise from the front of each cerebral ganglion. On either side of the front of the brain below the nerves to the antennz arises the slender pharyngeal nerve trunk, which shortly divides into three. The outermost of these three branches divides into two, one filament supplying the depressor muscle of the pharynx which arises from the postero-superior process, andthe other running: down inside the pharynx in close relation to its chitinous wall. The middle division of the pharyngeal nerve joins its fellow of the opposite side on the wall of the cesophagus as the latter enters the pharynx, the common trunk thus formed splitting into four branches to the intrinsic muscles of the pharynx. The innermost branch of each pharyngeal nerve joins a slender nerve arising in the middle line. The nerve thus formed supplies the pharyngeal muscles, but was not traced in detail. The brain is connected with the thoracic ganglion by the connectives, between which passes the cesophagus and which join after this to form a long 1906.] . The Internal Anatomy of Stomoxys. 527 connecting band as in Glossina. The thoracic ganglion is somewhat pear- shaped, and is supported by the internal chitinous skeleton of the thorax, from the surfaces of which arise the wing and leg muscles. Six pairs of nerves arise from the thoracic ganglion and supply the thoracic muscles. The abdominal nerve trunk continues from the posterior part of the ganglion running down in contact with the abdominal wall. It gives off three fine branches which supply the abdominal muscles, and ends in the third segment of the abdomen by dividing into three. Each of these branches again divides to supply the generative organs, the outer two running to the ovaries or testes and the middle one to the muscles of the ovipusitor or penis. Circulatory System. This consists of the heart and its continuation, the thoracic aorta. The heart is a tubular organ of the same type as in Glossina with chambers, ostia, and alary muscles. The wall, too,is composed of similar giant cells. Though several. stained preparations were made it was impossible, owing to the fat- body which obscured all detail, to count the chambers and cells in the heart wall. They seemed, from a comparison of all the preparations, to be reduced in proportion to the smaller number (four) of abdominal segments possessed by Stomoxys. eat The dorsal aorta consists of paired cells, as in Glossina, and runs up on the dorsal surface of the intestine to end on the cesophagus in a similar mass of cells. On the surface of the proventriculus, to which it is bound down, it becomes expanded and flattened, narrowing again to its termination. Male Generative Organs. These are comparatively simple. The testes (fig. 2, T.) are a pair of smooth, oval, orange-brown bodies with a shallow equatorial constriction. Their colour is due to a pigmented coat as in Glossina, but there is apparently not the same tubular structure. From the lower end of each testis arises a very fine duct (D.), short and straight, which runs down to join the duct of the opposite side as the upper limbs of a Y. From this junction an exceedingly short length of common duct (C.D.) runs into the bulbous upper end of a tubular organ, which would seem to function asa vesicula seminalis. ' This vesicula seminalis (V.S.) is a flexible tube, often. lying with two U-shaped bends in its course. At its upper part it is bulbous, gradually narrowing below this to end as an ejaculatory duct, which crosses the rectum dorsally from left to right, to enter the penis in front of it; it does not thus 528 Lieut. F. Tulloch. [Feb. 2, encircle the rectum as in Glossina. The hypopygium and penis are of the same type as in Glossina. x 24. Fic, 2.—Male Generative Organs of Stomoxys. Alimentary canal dissected out to show Malpighian tubes. Dorsal view. S.G., salivary gland; O., origin of Mal- pighian tubes; M.T., Malpighian tubules; T.T., thickened terminations of the tubules of one side; R., rectum; R.P., rectal papille, three of the four are seen through the transparent rectal wall; T., testis; D., duct of testis ; C.D., common duct ; V.S., vesicula seminalis. Female Generative Organs. The sex of a Stomoxys can be easily ascertained by inspection of the hind end of the abdomen; but, unlike Glossina, the scutellar bristles are of the same length in both sexes. The female generative organs are of the house-fly type. There are two ovaries (fig. 3, O., and fig. 4), each consisting 1906. ] The Internal Anatomy of Stomosxys. 529 of some 60 ovarioles. The ovary is moored to the body wall by a profusely branching trachea, which arises from the pleural space and ramifies among the ovarioles. In the natural position the ovaries lie with the long axis of the ovarioles pointing upwards towards the dorsal surface. Lach ovariole contains never more than four ova in various stages of development. The ovaries vary in size according to the degree of maturity of the lowest ova. In some flies they occupy more than half of the whole abdominal space. The ovarioles open into a wide tubular duct which joins its fellow from the other ovary like the upper limbs of a Y. Asa result of this junction is formed the common oviduet (fig. 3, C.0.), which runs down, forming a long third limb to the Y. Below the attachment of the uterine appendages the oviduct continues as the uterus. The appendages consist of the uterine glands and the receptacula seminis. The uterine glands (fig. 3, U.G.) are two rather stout tubular organs with slightly bulbous extremities. The bulbous end is firmly joined to the lateral oviduct by a very short double strand of connective tissue. Each gland ends in a short fine duct, and these ducts enter separately the shallow constriction which forms the arbitrary division between the oviduct and uterus. The receptacula seminis (fig. 3, R.S.) are two small, black, spherical bodies, each with a cellular socket resembling the fitting of an acorn cup. From this runs a very fine duct which enters the division between the oviduct and uterus in the mid-dorsal line. The receptacula are attached to each other, but can be separated by dissection. The distal portions of the two ducts are quite separate, but later each duct enlarges slightly, and from this point on to its insertion is closely attached to its fellow. This portion can, however, be separated by dissection, and it is then seen that the ducts are distinct and enter separately. The uterus (fig. 3, U.) is a tube of the same diameter as the common oviduct above, and runs down in the middle line into the ovipositor. The ovipositor (fig. 3, O.P., and fig. 4) consists of three cylindrical segments of thin chitin, which usually lie telescoped inside the abdomen. There is also a single external flap of dark chitin which lies folded up on the ventral surface of the fly. When the ovipositor is extruded by squeezing the fiy’s abdomen, the receptacula and uterus are pulled down with it, and can be seen through the transparent walls. : The upper segment of the ovipositor has three narrow ribs of dark chitin in its long axis, two dorsal and one ventral. The next segment is similar. The last segment has two dorsal plates only. The external flap, which is probably the third rib of the last segment, is, roughly, quadrilateral, and has two divergent prong-like processes arising from its free border. 580° The Internal: Anatomy of Stomoxys. L00e Ar R. Fig.. 3. x 24. te) % Fie. 4. Fic. 5. Fre, 3.-—-Female Generative Organs. The natural position of the parts has been consider- ably altered for the sake of clearness. Dorsal view. O., ovary ; R.O., right ‘ ‘oviduct ; C/O:, common oviduct ; U., uterus; O.P., ovipositor; U.G., uterine nyuni ©: )) gland ; R.S., receptacula seminis; R., terminal position of rectum cut short above. Fic. 4.—Mature ovary from another specimen. F Fig. 5.—Ovipositor extended. Dorsal view. The ventral ribs of chitin in the two upper segments are not shown. Specificity of the Opsonic Substances in the Blood Serum. 5381 ‘The points of the chitinous ribs which strengthen the segments of the Ovipositor project above the upper border of the segment, and to them are attached the muscles of the ovipositor. The narrowed terminal portion of the rectum enters the ovipositor on the dorsal surface of the uterus and runs down to the anal opening between the external plate and the last segment. The Specificity of the Opsonic Substances in the Blood, Serum. By WiiiAm Buttocu, M.D., and G. T. WESTERN, M.A., M.B. (Communicated by Leonard Hill, F.R.S. Received February 15,—Read March 1, 1906.) (From. the Bacteriological Laboratory, London Hospital, E.) A relatively high degree of specificity has been demonstrated for most of the antibodies which exist in immune sera, ¢.g.,in the case of agglutinins, lysins, preecipitins, antitoxins. With normal sera the proof of specificity is often difficult on account of the fact that the antibodies are present in the majority of cases only in small quantities. The following experiments are concerned with the specificity of the opsonic substances of normal and immune sera. As is well known, these opsonic substances, discovered by Wright and Douglas, act on bacteria in such a way that the latter become an easy prey to the phagocytic leucocytes. Ifa given serum be tested it will be found to exert an opsonic action on more than one kind of bacterium, and the question we have sought to answer is whether there is one or more than one opsonic substance; in other words, whether the opsonins are specific for the different bacteria on which they exert their opsonic action. In a previous communication* one of us (B.) has shown that when a microbe, ¢.9., staphylococeus, is digested with normal serum at 37° C. for 15 minutes, and the cocci are then brought down by the aid of a centrifuge, the supernatant liquid is found to be devoid of opsonie action for staphylo- cocci. Where the contact of the microbe with serum has been sufficiently long, and the centrifugalisation has been complete, the opsonin for the particular microbe is totally removed. * “Roy. Soc. Proc.,’ vol. 74. 532 Dr. W. Bulloch and Mr. G. T. Western. _—_[Feb. 15, We have attempted to determine whether the opsonins are specific by experiments of two kinds :— 1. The first method consisted in estimating the opsonic content of a given serum towards two different bacteria. A suspension of one of these bacteria was digested with the serum, and the mixture was thereafter centrifugalised, the resulting supernatant liquid being tested on both kinds of bacteria. To a quantity of the supernatant liquid the second bacterial suspension was added, and after the lapse of a certain time the centrifuge was again applied, and the resulting liquid was again tested. 2. The second method consisted in estimating from day to day the opsonic content of the serum of human beings suffering from lupus. At certain periods tubercle or staphylococcus vaccines were inoculated, and the effect on the two opsonic curves was determined. 1. Experiment on the opsonic action of normal human serum towards Staphylococcus aureus and Bactervum pyocyaneum respectively. Normal human serum was mixed with an equal volume of a suspension of Staphylococcus awreus, and the mixture was placed in the incubator for 1 hour at 37°C. At the end of this time the mixture was centrifugalised, the supernatant liquid “A” being removed from the deposit of cocci by means of a pipette. The supernatant liquid was in part retained, the remainder being digested for 1 hour at 37° C. with a suspension of Bacterium pyocyaneum, the latter being finally brought down as a deposit in the centri- fuge, leaving a supernatant liquid “B,” which was pipetted off. Result. 1. Normal serum (1 in 2 dilution)+staphylococci +leucocytes = 22°9) . 2 55 » (lind ,, )+2B. pyocyanewm+ 3 = 407 2.9 3. ” ” (1 in 4 a a ” at ” ool E = 4, Fluid “A” + staphylococcus + a = 05182 5 Be SOI? + B. pyocyaneum + » = +0| a x Cae eB sts » ie ” = O04 The contact of the serum with staphylococcus leaves the opsonic action of the serum for Bacterium pyocyaneum practically unchanged, the pyocyanic opsonin being finally removed by contact of the serum with this microbe. A similar result was obtained when the serum was brought to act on staphylococcus and tubercle -bacillus, as- can be seen in the following experiment. 1. Normal human serum was mixed with an equal quantity of an emulsion of tubercle bacilli in 0°85 per cent. NaCl solution. The mixture was digested for 30’ at 37° C. and then centrifuged. In this way a deposit and a super- natant liquid “ A” was obtained. 533 of Opsonic Substances in Blood Serum. uty O 1906.] Specific 80:0 90: 0 00: 0 00-0 00: 0 OT- 0 St: O OF- 0 08: 0 06. T 00: &1.0 03:9 OF. T 00: 4 19: T “TT 9dxgy eL-0 — €1-0 POO OCB (ease a ss +( “ ) snoosovopkyduys+( “ ) 4 980 “ ‘OT 80: 0 80: 0 0.0 srererereee (gatad g) soyAooong] + (yaud T) "GL + (syaud g) yu00 aod ¢g.0 ‘ourrrg ‘eT G0. 0 0-0 Sef RR eae CSRS BE) ete ttn (58S cel STR Sse ae) bre CD 0-0 0-0 0-0 ( = ‘ DE ee 2 ee ge SEA Ses JE HS Geel) AC) ve 0-0 0-0 0:0 (snosov0;Aydujs us Cee) WG aie SSS) Sa a8) 2) rere ae ery 60: 0 0-0 60: 0 srrereeres Corry, doy poureys) (syed g) voyMooone] + (quvd 7) ourpus + (syaud g) .y,, “ ‘TL 19-0 94-0 OBO ase (SE a) es ie wae) vine 2) nil “1 61-0 92-0 S1s0- = Ce ) SS See S) ra #i(Cesaes) wou 26 PE: 0 92-0 GF 0 POSIT (i, Se) < +( “ ) snovov0pfyduyjs+( “ ) Fie a. 8 1 = Demme ina eee ae eat Gee) fama >) gil 96: 6 8: OL OG PC aC tanicigeaty paisa (8) Vom 2 20 oF 0 g-0 PSO ier: leet a on B) ees ee ee ees) fecuraat( «Von PMU *S @ IL ¥- 11 OSE ie ae) se +( “ ) snodsosopkyduys+( “ +) g oures+[ wuutos peuoN “Pp FT rT rT saris ae (Se om) * re FY) ‘aL+( “ +) g oures+ 7 umaos poutony *g CF. ZI &- 21 Gisele (ee) s +( “ ) snooovojdydujs+( “ ) ‘wy ouryus + wns0s [eULAONT °Z ClO. € 0-€ £0. 2 “+ (sqred g) soyMooone] + (q1ed [) "A'L+ (810d g) -v:n ouryes + unaes [BULIONT “T “URE "M ‘T ‘1 3dxq ‘ayAoooneyT ted seqowolpy Jo ‘ON 534 | Dr. W. Bulloch and Mr. G. T. Western.” [Feb. 15, . Normal human serum was mixed with an equal quantity of an emulsion of es ylococcus aureus in 0°85 per cent. NaCl solution. The mixture was digested for 30’ at 37° C. and then centrifuged, a supernatant liquid “B” being obtained. | 3. The fluid “A” was mixed with an equal quantity of an emulsion of Staphylococcus awreus. The mixture was digested for 30’ at 37° C. and a deposit separated from a fluid “C” by the centrifuge. 4. The fluid “B” was mixed with an equal quantity of an enulen of tubercle bacilli. The mixture was digested for 30’ at 37° C., anda deposit separated from a fluid “D” by the centrifuge. : eh The opsonic content of the serum and of the fluids “A” «B” «