Materials for the Study of Variation Treated with Especial Regard to Discontinuity in the Origin of Species

I tesmarest's observat ion stood apparently alone until lately, when the specimen jusl described and several others presenting the same or -imilar variations were observed by I'.KNiiAM. Mr Benham was kind enough to ^ ml me the following specimens for examination: one female having a xiii<ilc r.rfrn oviducal opening on the left side upon the penultimate thoracic leg (Fig. 2(JC), and two females having a similar extra opening in the same place on the right ( Fig. :><i. ]i); in both of bhese the normal oviducal openings were unchanged. Together with these Mr llenham also sent a female haxing in, It/ ,,,!<• oviducal opening on the right side and another having ,,„/,/ tinii-ft oviducal opening (Fig. '20. A ), the correspondin^ leg of the other >ide having no trace of an opening.

ii'-ti- t.i Deemarett'fl jopri >h, «- that tli.' (lisrnptiun referred to a single specimen only. *85. After receiving these specimens I made an attempt to ascertain the degree of frequency with which such variations occur in FIG. 20. Females of Astacus fluviatilis having au abnormal number of oviducal openings. N.B. The form with three pairs of openings is not figured. A. Eight oviducal opening absent. B. Extra opening in right penultimate leg. C. Extra opening in left penultimate leg.

the Crayfish, and though the total number examined is too small to give a percentage of much value it may be well to record the result. In all, 586 female A. fluviatilis have been examined : of these 563 were normal in respect of the number of oviducal openings, and 23 were abnormal, as follows : 1 Mr B. ASSHETON sends me word of a similar specimen found among 80 of both sexes; Prof. W. B. HOWES of another among 144 of both sexes. In all ca-es of supernumerary oviducal openings the normal openin-- were also present.

These cases anin addition to those received from Mr Benhain. S.> far, then-ton-, the cases of extra opening amount to over 3 per cent, of females examined. Of 714 in<il<>x examined, only one was abnormal, having no tractof a generative op.-ning on the right side, the vas deferens ending lilindly and hanging free in the thoracic cavity. There was no fnnalr opt-ning in this specimen, and the abdominal appendag.-s had tinform characteristic of the male on both sides. The base of the last thoracic leg on the right side bore no enlargement for lipgenital opening, but was plain and like that of the penultimate leg1.

In cases of ft -males which lacked one of the openings, the basal joint for tli«- leg \vhidi should have been dilated and perforated tor the opening, was undiluted and resembled the basal joint of a penultimate leg. The oviduct upon the imperforate side was more or less aborted and hung loosely in the thoracic cavity. In the abnormal females with extra oviducal openings, the oviduct divides generally into two just before it enters the legs, the fork being placed at the level between them. In some few c ises no branch of the oviduct could be traced to the extra opening. In one specimen the extra opening led into a short tube \\hich ended blindly, not communicating with the oviduct. The specimen (4) with extra openings on the penultimate and last legs had thus in all six oviducal openings. Those in the normal position on the antepenultimate legs were of normal size, those on the next pair were smaller but still of fair size, while those on the last pair of thoracic legs were very small, that on the left side being the >malle.-t and admitting only a fair-sized bristle. In this specimen the single oviduct of each side forked in its peripheral third, giving a duct to each of the first two pairs of openings, but I failed to find any connexion between it and the openings on the last thoracic It-g^, which were very short blind sacs.

In all cases of extra oviducal opening the basal joint of the leg is expanded like those of the normal antepenultimate legs, the degree of expansion being proportional to the size of the opening. The normal openings are always the largest, but the extra ones are sometimes almost as large and would easily allow the passage of ova, but occasionally they are too small to let an egg through. As regards principles of Homoeotic Variation illustrated by these ca>es, three points should be especially remarked:

1 Compare the following : Astacus fluviatilis. Amongst 1500 specimens 3 were fmiiul in \\hk-h tlutuhfii-lo through which the green gland opens was entirely absi-nt. Tin- n]u niiiK' itself was not formed and the green gland of the same side was iil.st-nt. In another specimen the opening was deformed, probably owing to some mutilation. In this and the previous cases the green gland of the other side considerably enlarged. STKAUL, C., MUllt'r'.* Archie fiir Aunt. it. Phys., 1859,

1. That this Variation may be bilaterally symmetrical, but that the evidence goes to shew that it is more often unilateral. 2. That there is a clear succession between the several oviducal openings, those of the antepenultimate legs being the largest, the penultimate the next, and those of the last legs the smallest. 3. That Homceosis may occur between segments which are not adjacent, as in the case of extra oviducal openings on the last thoracic legs, none being formed on the penultimate (No. 84).

With the foregoing, the following evidence may be compared, though it is very doubtful whether it properly belongs here1. 86. Cheraps preissii [an Australian freshwater Crayfish, nearly allied to Astacvis]. Of seven specimens received one was a normal male and three were normal females. The other three had on the basal joint of the third [antepenultimate] pair of legs a round opening, having the size and shape and situation of the normal female openings. These apertures were closed with soft substance. The fifth legs bore the usual male openings, from which the ends of the ductus ejaculatorius protruded. The coiled spermatic ducts were normal ; but no ovary was found and no internal structure was connected witli these female openings. VON MARTENS, E., Sitzb. Ges. naturf. Fr. £erlit>, 1870, p. 1.

87. Astacus pilimanus </, a single specimen, and A. braziliensis J, a specimen collected by HENSEL in Southern Brazil, a similar opening was found on. the third pair of legs ; but in other specimens of these forms there was only a slight though sharply denned depression in the chitiiious covering at this point. VON MARTENS, E., I.e. 1 See also NICHOLLS, R.,PhiI. Trans., 1730, xxxvi. p. 290, figs. 3 and 4 describing a Lobster (Homarus vulgaris) having male organs on the left side and female organs on the right.

Though fr.'in the circumstance mentioned at the beginning of Chapter n, that the total number of segments in the Annelids is generally indefinite true Men-tic Variation cannot be easily recognized in this gn>ii]>, there is nevertheless a remarkable group of cases "C I'/iijH'rfect segmentation, in which by reason of the incompleteness <>t' the process of Division, the occurrence of Variation is at once perceived. The following cases were all originally described by CORI, who s] teaks of them as instances of "intercalation" of

meiits. For reasons sufficiently explained in the Chapter on V.-rt'-l.r.-e, there are objections to the use of this term, if only as a mod.- of expression, and the evidence concerning these cases has *ss. Lumbricus terrestris : the 4>0th segment having the form shewn in Fig. 21, I. being normal on the right side, but double on the left. Internally a septum divided the two parts a and a from each other. Ka.-h of them contained a nephridium, setce, &c. CORI, C. J., /./. w. Z., LIV. i,s!>2, p. :,71, fig. 1.

*Mi. Specimen having, in the region close behind the clitellum, three consecutive segments, each resembling that just described. Of the>e the first was double <m the right side, the second on the left, and the third on the right, again. Fig. 21, II. shews the internal struct nre. the nephridia and other parts having doubled in each of the doubled half-segments. CORI, I.e., p. 572, fig. 2. !)0. Lumbriconereis : case similar to the first case in Lumbricus,

HI. Halla parthenopeia. A specimen 50 cm. long presented numerous al.iiorinalit ies of which two are represented in Fig. 21, IV. At the point marked ,/' the lines of division between the segments enclose a small spindle-shaped island of tissue. Three segments lower a wedge-shaped half-segment is similarly formed. At FIG. 21. Examples of imperfect segmentation in Annelids (after GOBI). I. Lumbricus terrestris (No. 88). II. L. terrestris (No. 89), as seen when laid open on the dorsal side. III. Lumbriconereis (No. 90). IV. Halla partlienopeia (No. 91).

N, nephridium ; Np, nephridial pores; D, alimentary canal; dG, dorsal vessel; vG, circular vessel. The letters a, b, c, &c. indicate the parts belonging to the respective segments. another point in the same animal (not shewn in Fig. 21) one of the segments was partly divided into two in the right dorsolateral region. CoRi, p. 572, figs. 8 and 9. 92. Lumbricus terrestris. Fig. 22, I. A shews a part of an Earthworm seen from the dorsal side, the ventral side being normal in appearance. By following the groove indicating the plane of the septum between b and c ori the right side to the ventral surface, it could be traced to the left side between b and c, so across the dorsal surface, between c and d on the right side, across the ventral surface and between c and d on the left, reaching nearly to the middle dorsal line again. This is shewn diagrammatically in Fig. 22, I. B.

93 A simpler case affecting one segment only is shewn in Fig. 94. Another specimen exhibited a similar arrangement near the tail-end (Fig. 22, III.). The lettering of the figure sufficiently explains the course of the spiral septal plane. [Cori does not state that the septa internally formed a spiral division, but it can scarcely be doubted that they did so, following the external groove, I'i •.. '_'-!. Sjiiral segmentation in Littnbricns terrestris.

II. A, the case No. (,(3 ; II, B, diagrammatic representation. Two oilier rases described by CORI may be mentioned here, though tin-re is ;i presumption that they are not really examples of Variation in the .segmentation along the axis of a Primary Symmetry, but rather In-long to the class of Secondary Symmetries. They are alluded to liere as it is convenient to illustrate this distinction by taking them in connexion with the examples just given.

'95. Hermodice carunculata. ( Fig. I'D, III.) Between two normal segments is what seems at first to be a segment double on the left MI!.- with two complete sets of parapodia, but imperfectly divided on the right (left of figure), the septal groove stopping short before it readies the parapodia] region. The lower half on this side is repn ented with a normal ventral ramus of ihe parapodium, Imt the particular two cirri Cv. I. and Cv. II. and two branches of setae. The eonilition of the dorsal ramus is not described. Of course without seeing this specimen it is impossible to say more than this, but the figure strongly suggests that the division between the two halves of thU parapodium was a i/irixion lut,, 'unity* and not into successive aenta. The lignre represents the lower cirrus Cv II. as standing in the normal position for the cirrus, on the posterior limb of the parapodium, hut the .-interior cirrus is distinctly shewn as placed on the anterior limb of the elevation and anterior to the bristles. If this were aetually the rase, this doiiNe parapodium must be looked <>" as a kind of l.ud. with a distim-t Se.-omlary Svmmetry of its own. he.M-rilicd afresh from Com, C. J., Z. /.' </•. Z., L1V. 1892, p. 574

96. Diopatra neapolitana. In the middle of a specimen 35 cm. long was an arrangement somewhat similar to the above. The part marked FIG. 23. I. The case of Diopatra neapolitana (No. 96) from the side. II, the same looking upon the parapodia. C I. C II, the two supernumerary cirri. III. The case of Hermodice caruncitlata No. 95. Cv, cirrus of ventral branch of parapodiurn ; Cd. dorsal cirrus ; Cv. I, Cv. II, the two cirri borne on the supernumerary parapodium. (After Com.)

b' was cut off as shewn in Fig. 23, I., it bore a normal cirrus, and the other part of the segment, marked 6, bore two cirri and two bunches of bristles. The figure does not indicate that there was any relation of images between these two parts, but this would scarcely appear in this case unless specially looked for. Described afresh from Com, C. J., I.e., p. 573, figs. 10 and 11. In considering the evidence as to Secondary Symmetries reference to these cases will again be made.

The number and ordinal positions of the primary and accessory generative organs and of their ducts differ in the several classificatory groups of Earthworms. In the evolution of these forms it may therefore be supposed that Variation in these respects has occurred. To this subject the following evidence relates. The difficulty which was mentioned in the case of Variation in vertebra?, that there is no clear distinction between Homoeotic and strictly Meristic Variation, will here also be met, inasmuch as the total number of segments in these forms is indeterminate ; but

1 For information and references on this subject I am indebted to Mr F. E. Beddard and Mr W. B. Benharn. probably we shall be right in regarding the majority of these variation- .1- Homujotic. Li M MKICUS. Throughout this genus there is normally a single pair of ovaries, placed in the 13th segment, on the posterior -u i lace of the septum between the 12th and 13th segments. The following cases of supernumerary ovaries are recorded : 97. Lumbricus turgidus : specimen having an extra pair of ovaries in tin- 14th1 segment.

98. Specimen having an extra ovary on the right side in the 14th 34 fluent. 99. L. purpureus : specimen having an extra ovary on the left M<1'- in the 1 4th segment. In all these cases the extra ovaries were in size, form and P"-ition like the normal ovaries. There was no extra oviduct or rec.'ptaciilum ovorum, but the normal ovaries and oviducts were present as usual. BERGH, R. S., Zeit. f. wiss. Zool, XLIV. 1886, p. 3< ),s, note. *10(). Allolobophora sp. [partly = Lumbricus, the common Earthworm] : specimen having, in all, seven pairs of ovaries; viz. a pair in the 12th, 13th, 14th, loth, 16th, 17th and 18th segments. Of tlu >e all except the pair of the 13th segment are supernumerary. Kach of these ovaries was placed on the posterior face of a septum in the n-iial position. The three anterior pairs in shape, structure ami portion closely resembled the normal structures. Of these the most anterior were slightly the largest. The four posterior pairs were smaller and resembled the ovaries of a very young or immature worm, but on examination all were found to contain "va. The normal pair of oviducts were present and no extra oviducts could lie found, though carefully sought for. WOODWARD, M. F., 1'. Z. 8., 1N!>2, p. 184, Plate xm.

*101. Lumbricus herculeus, Savigny (= L. ayricola, Hoffmeister), having an asymmetrical arrangement of the generative organs, &c. < >n the left side the arrangement was normal ; the ovary being in the 13th segment, the oviducal opening in the 14th, and the opening of the vas deferens in the loth segment (Fig. 24). < >n the ni//if side each of these structures was placed in the segment anterior to that in which it is normally found: the right ovary was in the 12th, the external opening of the right oviduct was in the 13th, and the external opening of the right vas deferens waa iii the I4th segment. The spermathecawere normal on the left side, being placed in the 9th and 10th segments, but on the right side one spermat heca only was present, that of the 9th se^mrnr. r|-],,. vesiciihe seminales were present as usual in the !>th and llth segments, but there was no vesicula in the 12th

In I'.i nun's enumeration tlie ordinal number of these segments is one less than in that commonly used: the latter sv.-t. m i> mln^ti'd above. segment on the right side, while that of the left side was fully FIG. 24. Lumbricus herculeus, having the generative organs of the right side one segment higher than usual. A, external view from below. B, view of the organs from above, spth, sperniathecfe ; or, ovary ; CBS, oesophagus ; cat, calciferous glands. After BENHAM.

formed. It is remarkable that in this case, the calciferous gland of the 12th segment was absent on the right side. [I am indebted to Mr Benham for an opportunity of examining this specimen.] BENHAM, W. B, Ann. & Mag. N. H., 1891, Ser. 6, vn. p. 257, PL in. 102. Another specimen presented the same variations as the foregoing, both as regards the asymmetrical arrangement of the genital pores and the absence of the calciferous gland : but in it there were vesicular seminales on the right side in segments 10 and 11. but none in segment 9 ; and there was a spermatheca on the right side in segments 8 and 9. [In the normal form the spermatheca? are in segments 9 and 10, so that, in this individual in the matter of the spermathecse as well as of the genital pores, structures were formed in particular segments which are normally found one segment lower down.] BENHAM, W. B., in litt., March, 1891.

103 T-ible shelving position of ovaries in forms having two or more pairs ,,t',.,;,nes,andin the Variations found (slightly altered from 104. Allolobophora >j>. [,/ pores normally in loth and $ pores in 14th, as in cuiiiiii.ni Earthworm]: specimen having on the right siili- £ pore in 20th and $ p..iv in 19th ; on the left side, £ pure in 17th and $ jii.n- in Kith. MiCHAELSEX, W., Juhrb. Jfnniltnrg. wiss. Anstcdt, IN!K), vn. p. s. In c-ach case the £ pore is in tinaegini-nt lifhind tin- $ pore, as normally. The position of ovaries n. .t

Lumbricus agricola Hofl'm. (= terrestris L.) : amongst 230 specimens in which the ] 'i '-it ion of the male pores are determined, 6 specimens were found in which these openings were not normally placed (vi/.. one mi c;ich side in the loth segment). In two of these specimens, linth puivs were iii the 14th segment; in one case the left pore w;i-. in ilie llth segment ;md tlie right was in the loth: these three \\iirms were ( leniiaii. One specimen was found in Savigny's collection in I'aris \\liich had two pores on the left side [and none on the right i'i . In <>ne Knglish specimen the " vulva" [.sr. the two male pi ires] was in (he 1 Itli segment and in another it was in the 16th. [The author speaks sometimes of both pores as the "vulva," and at ..I her times he uses this term for one pore only, but the meaning is plainly that given above.] HoFKMKisTKH, W., Uebersicht alter bis j.i-.i /„ /.-,!,,,,/,, i Art' it a, if. /•'"//////•' Regenwurmer, Braunschw., ~

*106. Perionyx excavatus. In this earthworm a very remarkable series of variations has been observed by Beddard. The accompanying table shews the varieties in number of spermatheca? and position of the generative openings which were found. The spermathecas are generally 4, and are placed in the 7th and 8th segment, but in several specimens there were 8 and their position varied from the 6th to the llth segment. In all the varieties, however, they were in segments adjacent to each other. In four specimens the spermathecse were in the 8th and 9th segment on the right side and in the 9th and 10th on the left. In normal specimens the male pores are 2, but individuals with 4 (and perhaps G) were found. There are generally 2 pairs of ovaries and oviducts. In Var. No. 11 an additional ovary was found on the right-hand side in the llth segment and in Var. No. 10 there were three pairs of ovaries.

Though the position of both varied greatly, the male pores were always posterior to the female ones. In some specimens certain of the segments were only divided from each other on one side of the body, being confluent on the •other. For example in Var. No. 14, segments 11 and 12 and also segments 18 and 19 were only divided from each other on the left .side (cp. Nos. 88—91). Out of 430 individuals 15 variations in these structures were seen; of 12 of these variations single specimens only were found, but two specimens occurred with each of the other three forms of variation. In a single case a nephridium was found nearer to the dorsal line in one segment than in the adjacent segment. Many of the conditions here occurring as variations are found normally