s 1: 37 ON THE STRUCTURE OF THE RESPIRATORY ORGANS OF THE TERRESTRIAL ISOPODA. By ERNEST E. UNWIN. Plates X.-XVII. (Read 13th July, 1931.) 1. INTRODUCTION. Terrestrial isopuds present problems in respiration be cause of their change of habitat. With organs of respiration which are homologous with the tbranchire of Asellus and other aquatic isopods, they live on land and breathe air. They are found in very varied situations; some require very wet conditions, others can live in comparatively dry places. In this communication I propose to describe the inter esting anatomical features of the respil'atory organs of certain species. In the species treated of, and :in the Oniscaddea generally, five pairs of ahdominal app·endages take some part in respiration. I •propose to describe these appendages in the female forms only. T·here are no differenc.es in the essential points of histology and physiology :between the male and female pleopods, and the fact that, in the male, reproductive processe·s are associated with the 1st and 2nd p1eopods, brings in an unnec·ess.ary complication. Methods. Living Specimens: I have indicated the methods in use for examining the abdominal appendages in situ. Abdominal appendages as flat objects: Methylene blue, as a stain for living tissue, as well as for fixed ones, is most useful. Prussian blue gelatine injection has .been used to investigate the limits of blood cavities. Paraffin serial sections: For fixing I have found Flem ing's solution the best. Hot 80 p·er cent. alcohol has given good re·sults with some. Staining: Delafield':s hrematoxylin followed by eosin in 90 per cent. alcohol. The staining has been done on the slide in dip.ping jars. ~L -- .. ·<' 38 RESPIRATORY ORGANS OF TERRESTRIAL ISOPODA p, and P. Roy. Soc. Tas., 1931. Plate X. II. SPECIES DEALT WITH. 1. Trichoniscu-S pusillus, Br. Pl. X., figs. 1-6. T.he ventral surface of the abdomen is shown in Fig. 1. It is covered by five pairs of oval plates, which overlap from before backwards, and those of either side meeting together T,ICH0Nl5C.US ru.rLLUS ~. in the mid-ventral line. Pr. Normally each of these plates indicates an abdominal ap.pendage, consisting of a basal part, an outer plate (exo.podite) and an inner plate (endopodite), but the endo podites are missing from the first two .pairs of appendages. Thus ther·e are five pairs of exopodites and three pairs of endopodites. These are the respiratory organs. In manY' forms the structure of all the exopodites is similar and the same applies to the endopodites; it will be convenient, ther·efore, to take the 3rd .a~bdominal appendage for spe·cial study as this is the first complete appendage. Where differences exist hetween tpe exopodites of the 1st and 2nd abdominal appendages and that of the 3rd, the descrip tion will be extended to include them. \"::· Figs. 2 and 3 show the 3rd abdominal appendage seen F;j· "\: from the ventral or outer, and from the dorsal or inner aspect. The exopodite acts as an operculum for its rather ~..t E:K. smaller and more delicate endopodite. Each plate takes .• · e.. . part in respiration. This has been made clear by observa tion, by experiment, and by the evidence afforded by sections. If a living specimen is obseryed under a low .power of the microscope, the abdominal appendages seem covered with a film of water, at least the endopodites are, for as the· exo podites are raised and lowered~an action which, although not absolutely regular, is fairly continuous when the creature is ·.. c. ....\ .. ~11 \/t.. not moving and which is a breathing action-a film of water t(U 4J retreats and advances under the free edges of the exo·podites. Trichoniscus pusilltts . ~ . This shows that the endopodites are kept very wet and are used for res.piration in a very damp condition. By using a higher power the circulation of blood can be observed in the exopodites. They are well supplied with rblood, and act as additional rbreathing organs. I proved this by putting .the endopodites out of action, in this and in other woodlice, bY' inserting some vaseline under the exopodites. The creatures ,. lived for very considerable periods. In the control experi- ~~ ~ ~/ BY ERNEST E. UNWIN 39 ment with the whole of the ventral surface of the abdomen made impervious, the woodlice did not live more than two days at the longest. It is also clear from the transverse sedions that the exopodites are used as organs for respiration, and that the outer or ventral surfaces function to this end. The chitinous cuticie of the ventral or outer wall-the wall towards the air, the oxygen medium-is very thin and the inner or dorsal wall quite thick. If the exopodite functioned as an oper culum solely, the outer or ventral wall would be the thicker, as it is for instance in the 3rd abdominal exopodite of Asellus. This wiU be referred to :below when the transverse sections are described in detail. We shall find as we stu·dy the woodlice adapted for drier situations, that the exopodites become more and more important as breathing oTgans. It is in the exopodites that accef.sory breathing organs are developed; and the endo podites, although three pairs are always present, become Jess important in consequence. The Exopodite: T.he exopodite is roughly oval in out line. (Fig. 2.) When seen under the microscope as a trans parent object, either in the fresh .Jiving condition or treated with methylene blue, the surface is covered with small dots, which are •not arranged evenly over the surface. These dots are the nuclei of the epithelial layer, and here and there where they are grouped together comes an inter-lamellar pillar or bridge (le pilier of Kim us (1)). The space between the two walls traversed by these .pillars is the blood cavity. There are no definite blood vessels, but at the sides of the appendage channels can be seen, free from pillars. These are the inhalent and exhalent blood channels, or afferent and efferent canals. It has already been noted that the circula tion of the blood can be seen. This is due to the· large size of the blood ·corpuscles. The blood collects into a large v·entral sinus in the abdomen and is drawn through the abdominal plates by the pumping action of the heart, which, receiving purified blood, propels it forward to the head and to the body generally. The blood enters the exopodite at the point of attachment, and Tuns along the inner border in what we may call the afferent canal, finds its way between the two lamellre, into the efferent canal along the outer border, and ·so out of the appendage. The blood is prevented from short circuiting from the afferent to the efferent in (1) Kimus (see Bibliography). v 40 RESPIRATORY ORGANS OF TERRESTRIAL ISOPODA BY ERNEST E. UNWIN 41 the proximal part of the exopod.ite iby the presence of a use when the air enters the space in which the endopodite tissue which is continuous with and. similar to the tissue lies. By the following extract it is clear that he imagine>< found in the body cavity. This is the intermediary tissue to only one kind of respiration--that taking ,place in the endo use the name suggested. by Kim us (le tissu intermediaire). pcdite cavity, whereas the outer 1surface of the exopodite is adapted for respiration of a different kind. (See Sum If the exopod.ite he examined. by transverse or longi tudinal section (Figs. 4 and 6), the details of the two walls mary.) or lamellre can be seen. Each lamella consists of an "It is true that as the blood circulates through the gill, epithelial part and. a cuticle, and the two lamellre are very •· it is separated from the external atmosphere by the rather different. The inner or dorsal wall ·has a fairly thick cuticle, " thick ventral wall of chitine. But through the movements and the epithelial lining is quite distinct with its nuclei at " of the gill, taking place as described for Porcellio, air is frequent intervals. This epithelial lining is a single layer " constantly brought into relation with the inner surface of cells which give rise to the cuticle on its outer face. The "where the chitinous wall is thin." (Stoller.) cuticle of the outer or ventral wall (or lamella) is very thin, It will be best to correct at once this serious mistake in and the epithelial lining is also extremely thin, indeed in Stoller's account, for it occurs in his descriptions of other most sections it is not seen except in the region of the pillars, species. In his description of the last three pairs of exo but at the time of the formation of a new cuticle the po·dites of Porcellid. scaber, which are more or less normal in layer is distinctly seen. In a later communication I have structure and do not possess the air-tree organs, he describes described and figured a section through an exopodite at the their function as protective covers for the endopodites and time when this layer shows best of all (Pl. XIII., fig. 28). their movement causing renewal of air to the ·endopodites. Sometimes the very thin remains of the layer is seen when by I agree in these two functions, but do not agr·ee with the some action (osmosis) the protoplasmic lining becomes dis cbs·crvati01n given below as to the structure of the wall. torted. The pillars pass from cuticle to cuticle. The outer wall of the exopodite is thus adapted for respiration through "It is pos·sible that the outer gills ( = ·exopodites) in "a.ddition to performing these functions, contitute in them its very thin wall; the inner wall for protection and for the necessary support of the exo.podite itself (Fig. 6). •· selves organs of respiration. The blood circnlates betwe.en " the two walls of the gills, and whether respiration takes Stoller (2), to whose work upon the hreathi,ng organs of " place depends upon the penetrability ;of the walls to the woodlice I shall have occasion to refer and to correct in many " gases to be .exchanged. The ventral wall of chitine is important particulars, does not describe Trichoniscus, but " thick-thicker than the .chitinous wall in some other regions refers very briefly to Ligidiwrn hypnorum. He says, referring "of the body as, for example, the thoradc legs (in which also to a cross section of the exopodite of which he gives a figure: = "the blood circulates). The dorsal wall is, on the other "It is seen that the gill ( exopodite) is in prindple a "hand, modeTately thin. It is, moreover, exposed to the •· simple sac containing blood. The outer layer of the wall, " moist air in the chamber in which the inner gills lie. It "·c·omposed of chi tine, is much thicker on the v·entral than " would seem probable, therefore, that to some extent " on the dorsal si·de of the gill. T·he hypoderm ( = "respiration may take place through the doTsal wall." "epithelium) everywhere lines the chitine, and in general (Stoller.) " varies with it in thickness." This extract goes to ·show that in Stoller's mind the This is the exact opposite of what I find in Trichoniscus, dorsal wall, the one that is applie·d to the endopodite, is and at first I thought that this section of Ligidium must have considered thinner than the ventral 'Wall, the one that is become turned. round in some way, or that we were using bathed by the external air. I have examined all the common the terms "dorsal and ventral" in different ways, but a few species of woodlice and fi!!d the ·oppo-site arrangement in lines further on he makes it quite dear that he looks upon every case (except where a great change Iike the develop the inner surface (that is the dorsal surface as the plates ment of an air-tree organ has altered the ex:opodite). In are found in situ) as the respiratory one, and that it is in Ligia oceanica, Trichoniscus pusillus and roseus, Oniscus. (2) Stoller (see Bibliography). asellus, Philoscia muscorum and conchii, Porcellio scabeT, ,. r"'~ 42 RESPIRATORY ORGANS OF TERRESTRIAL ISOPODA BY ERNEST E. UNWIN 43 Cylisticus convexus, A rmadillidium vulgar·e, Porcellionides one wall unite with the cells of the opposite wall. The pr"ldnosus, Platyarthrus hoffmannseggii, the ventral wall is question of these pillars is discussed in the summary. the thinner, and is normally adapted for respiration. We shall see that some species use this ventral surface more than 2. Oniscus asellus, Linn. others, but there is not the slightest doubt hut that Stoller PI. XI., figs. 7-16. is incorrect in his description. A glance at Figs, 4, 25, 53, 57, 59, and 69 will show this. The general arrangement and number of the abdominal appendages are the same as .in Trichoniscus pusillus. Look The Endopodite: It has already been pointed out that ing at these appendages in situ, a peculiar feature is at once there are only thre·e pairs of endo.podites. The 3rd abdominal seen. The outer border of each exopodite has quite a differ exopodite is the first to have a corresponding endo·podite. ent appearance from the rest of the exopodite. This outer This latter is slightly smaller in area than its exopodite, but border appears as a silvery white segment, rather less than its structure is in marked contrast, as can be seen in figures a semi-circle in shape and occupying from one-fifth to a 3, 4, and 5. The chitinous cuticle is very thin, and the quarter of the area of the whole exopodite. (Pl. XI., fig. 7.) epithelium is, on the contra~y, very thick, and the whole This unusual appearance attracted my notice, and I began appendage is spongy. There is one peculiarity which I have to study it by observations on .Jiving specimens and by means not noticed in the endopodite of any other woodlouse, and that of sections. W·hen watching the breathing action of Oniscus is a lobe which is seen attached to the inner distal corner of under a low power of the microscope, it was clear that ·t;hese the appendage. The blood cavity of the endopodite is con bord·er organs had bubbles of air underneath them, and that tinuous with that of this lobe. In Fig. 3 I have attempted to the silvery appearance was due to this. With a little patience sho'\v the nuclei and the large epithelial ceUs as they are one could get the air replaced by water, and then the silvery seen after the endopodite ihas been treated with methylene ap.pearance disappeared. By flooding the Oniscus with weak blue. In Fig. 5 a portion of an endopodite is shown as seen alcohol, the air was dislodged, and the alcohol could be re in cross section. The large nuclei are very clear, but the placed by water before any p.ermanent injury was done. In cell boundaries cannot be seen. There are very few pillars this way I' satisfie-d mysdf that the silvery appearance was running frorri the epithelium of one wall to that of the other. entirely due to the air imprisoned under the corrugated They are narrow pillars, and as one follows them into the cresc·entic lobes. I mention this here because below I shall epithelium radiating fibres can he S·een. The bloo·d cavity is refer to Stoller's work which later caused me to revh:!w the very variable in size, in some sections the epithelial linings are c.Josely opposed, in others they are widely separated. whole matter. These ·observations support Kimus's suggestion that the This special part is very thin and, having its outer or function of the .pillars is partly to assist the flow of blood ventral wall flus·h with that of the ordinary part, there is a through the appendag·es. considerable space hetween its inner or dorsal wall, and the und·erlying exo.podite or the body wall. This is the air or The pillars are very similar to the type found in the lung cavity, and the air is renewed :by the raising and lower endopodites of Asellus aquaticus, and to which Kimus has ing of the exopodites as weU as by the· flexing of the abdomen devoted so much attention. The chief point to note here is and other movements caused by locomotion. that these pillars are not complete cells, but formed of the fused p1·otuberances which are carried· by the large epithelial Besides this special arrangement, designed, as we shall cells. Kimus in summarising the kinds of pillars found says se·e, for more efficient respiration of air, Oniscus has exo of this kind- podites and endopodites which, similar to those already a descrihed in Trichoniscus, are adapted for breathing of moist "Enfin, le pilier forme par !'union, travers la cavite a air. The endopodites, although 1not so wet as those of Tri "lbranchiale, de protuberances .appartenant deux larges choniscus, are yet obviously meant for the same kind of "cellules epitheliales, qui en possedent plusieurs semblables." l'·espiration, and the habitat confirms this .conclusion: for The pillars of the exopodite are quite diff.erent. They Oniscus is oft-en found along with Trichoniscus in very damp are formed of the union of two or more cells. The cells of situations; but the presence of other :breathing organs en- E ¥' 44 RESPIRATORY ORGANS OF TERRESTRIAL ISOPODA P. and P Roy. Ho~. Tas., 1931. Plate XI. abies it to exist in places which would ibe too dry for Tri chaniscus. I shall refer to this in detail later. The 3rd Abdominal Appendage: Fig. 8 gives a general view of the appendage from the ventral side, and in Fig. 7 the 3rd exopo-dite on the right side has been turned ba·ck 0N15C.US ASE:LLUS. '(.,. to show its corresponding endopodite. Reference to these t•. drawings will save much d-escription. T.he ordinary part of the exopodite and the endopodite exhibit much the same structure as those appendages in Trichoniscus, although the .1 ~ larger size makes it easier to e;xamine the histalogy in detail. . . Reference to Figs. 11, 15, 16, especially to Fig. 16, which is a -,"en longitudinal section passing through an exopodite and its ·.3"'f>< h: ,''.:.J · QI ' mdopodite in such a way as to pass through the ordinary part of the exopodite, will show the chief points of interest. The Exopodite: In the main part of this .plate the inner or dorsal wall has a very thick cuticle, carrying still further d..r. --.-.. - the arrangement we have noted in Trichoniscus, and to which we have already referred at some length. The outer or lr.c_. ··E.x ventral wall is very thin, and points to the use of the outer face of the exopodite as a 'V'aluable .breathing organ, for not only is the chitinous cuticle very thin, but the epithelial layer is also very much reduced, so much so that in places o..j{, it is difficult to make out. Thus the blood which flows through the appendage is brought into close relationship ... with the air outside, and the necessary interchange is pos·sible. F"l'j· 12.. :(,_..._t.. ·.~.,r""· <:. ..... ~ ~; Fig. 13. The pillars are more frequent here than in Trichoniscm. c~. They resemble the simpler pillars found in Asellus, described by Kimus i•n his memoir. At the base of the appendage the third tissue, the intermediary tissue, can be seen, pushing a little way into the appendage acting as a partition to direct the flow of blood and to prevent the blood staying in the d.l.lc proximal part of the appendage. This tissue, consisting of ~- elongated stringy cells with prominent nuclei, is seen in sec [JC. '!- -~.e. tions, and can also be ooen in the appendage after treatment Fi'l It ""-~·· \V..c with methylene blue, and examined as a flat object. In the special ;part of the exopodite, this tissue becomes a very thin threadlike partition, sometimes double, continuous with a similar partition in the proximal part of the appendage, P· f•'j 16 especially in the part abutting on the special part. This t~7 E·tu·•·l, partition serves to divide the internal blood cavity into an Oniscus asellus ~ . upper and a lower cavity. The structure of this special part of the exopodite must be dealt with in -some detail, for it is here that I find myself in serious disagreement with Stoller. After reading his ---- -; r BY ERNEST E. UNWIN 45 account I repeated my own experiments and sectionised the appendages again, and also tried the experiments suggested by Stoller. I will first of all give a detaile-d description of this part as it is seen by me, and then give Stoller's view, pointing out where I think he is in error. The figures make clear the general structure. We notice that the exopodite as it approached the outer border where this special segment is situated becomes very much thicker, and then suddenly narrows to less than one-third its thick ness; the dorsal wall is the one that alters, the ventral wall of the special part and that of the rest of the exopodite are flush with one another. Besides the change in levels the inner or dorsal wall changes in the thickness of the chitinous cuticle. In the ordinary part of the exopodite it is very thick, in the special part it is exceedingly thin, the ,thinnest cuticle found anywhere in these appendages. The outer or ventral layer of cuticle is very thin, but decidedly thicker than this on the inner surface of the special part. This points to the fact that the inner surface is of more importance from a respiratory point of view. Both surfaces are thrown into corrugations, but the'Se are much d·eeper on the inner surface. Turning •now to the internal structur·e. In all sections, transverse or longitudinal. a tpartition is seen running through about mid-way between the two faces, but the deeper corrugations of the inner .or dorsal surface make the cavities on that side rather larger. This partition consists of elongated threadlike cells with elongated nuclei. The partition in some places is double with a very narrow space ibetween the two threads, but in other places it appears as a single partition. This partition, the intermediary tissue, is continued into the ordinary part of the exopodite for a greater or lesser dis tance. Near the proximal portion ·Of the exopo·dite this partition can be traced into the ·protopodite, but in the distal part it can be traced but a short way. The internal ·cavities are interrupted by the pillars which pass from one face to the other. It is usually only at the base of these pillar eells that any trace of the epithelial lining can be made out. This lining is extremely thin. The pillars are of a somewhat different nature here in that they are grouped together into bands. This arrangement, together with the intermediary tissue, divides the cavity into two sets of radial passages, one set of passages on each side of the partition. This can be seen in a surfac.e view, and an _L r-;?"'" i 46 RESPIRATORY ORGANS OF TERRESTRIAL ISOPODA BY ERNEST E. UNWIN 47 attempt to show the radial arrangement of pillars is seen "In the general part of the gill there are two of these in Figs. 8 and 9. If a living specimen is examined under a "chambers, an inner or dorsal chamber and an outer or low power, it is easy to observe the course of the blood flow " ventral one. They contain blo.od. Each chamber is ing through these radial passages. The blood enters the " bounded above and below by a thin line. It results from appendage at its place of art~culation, which is near the outer "this that the cavities of the two chambers are separated by border. The blood stream divides, some of the blood going " a double wall. This appears as two fine parallel li•nes be to the exopodite in the ordinary way passing down the " tween which is a. very narrow space. Lying in this space afferent canal along the inner border and working its way "and occurring at frequent intervals are elongated nuclei. across the appendage into the efferent canal, which is the " The outer walls of each chamber lie immediately within boundary between the two parts of the exopodite, and, by "the hypodermic (e.pithelial) layers on the respective dorsal this, out of the appendage: other blood takes a different "and ventral sides of the gill. The outer walls have nuclei, course, it goes to the spe·cial part. The transparent character "but they occur at widely :separated intervals. The hy.po of this special part makes it comparatively easy to observe " dermic pillars cross the blood chambers. Where they occur the circulation. It is be·st to remove the last pair of legs, and ,; the walls of the chambers curve round them, forming then invert the creature in a watchglass of water, and rest a " sheaths. Thus the cavities containing blood are every cover glass on the .creature to restrain its movements. The " where separated from the hypoderm by a boundary wall.'~ blood corpuscles travel through the ventral radial passages (the uppermost ones, as the animal is .placed) towards the Here are several inaccuracies :-The main portion of the margin of the sp·ecial part. The partition is so transparent exopodite is certainly a simple cavity crossed by pillars lined that it is not visible, but its presence is obvious, for as the by epithelium. thicker ·on the· dorsal than on the ventral side corpuscles get to the margin they suddenly fall through, as it (Fig. 16). In the proximal part of the eX"opodite and in were, to the lower chamber, and one can see them travel back, neighbourho·od of the spE:cial part this thin partition is pre slightly out ·of focus, in the dorsal radia.l passages. The sent, but not in every part of the exopodite. The nature of partition comes to an end just before the margin is reached, this partition; its simplicity in structure to the tissue in the which accounts for the obs·ervation above. Thus in Oniscus body cavity; its presence in the protopodite and in the proxi we have, besides the main portion of each exopodite, which is mal portion of the appendage as a blood directing tissue; similar in function to the exopodites of Trichoniscus with its presence in the basal ,part of the endopodite as well; and its outer or ventral surface adapted for respiration, a set the fad that a similar tissue is found in Asellus and other of special breathing organs developed on the outer border aquatic isopods are points with which Stoller does mot ·II of each exopodite. Enough has been said to show that these seem famHiar. Then he is mistaken when he mentions that are most valuable additions to the respiratory powers of this the blood cavities have definite boundary walls, that is of 11;! creature. The blood in the passages is only separated from cellular structure, other than the epithelial lining and in some the air outside 1by the thinnest of walls, and the constant cases the intermediary tissue. He, himself, is evidently in supply of blood and the repeated change of air supply the doubt as to this, for in a footnote on page 14, added after requisite conditions for an efficient breathing organ. the treatise was written, he leaves the question undecided. The account given above differs in several important The footnote is meant to qualify the following statement:- II;; points from that furnished by Stoller. It will be best to "The blood cavity occupies the whole space within the quote from his treatise the paragraphs with which I am in disagreement. "hypodermi·c layer, and is bounded by a very thin wail lying "contiguous to the hypoderm :and conforming to· its irre P. 25-" ... In Oniscus the general cavity of the gillx " gularities. In this wall at long intervals lie elongated " is not a simple cavity as in Po1·cellio, but is separated into " nuclei. I infer from the relations of this wall that it is "a number of chambers ·by means of partitions lying parallel "of mesodermic origin." (Stoller.) " to the walls of the gill. (xGiU or outer gill ·is used for exopodite; inner gill for In the footnote he refers to the work of Leichmann, who endopodite.) has describe•d the brood pouch lamellre of Asellus aquaticu.s, '' '¥ 48 RESPIRATORY ORGANS OF TERRESTRIAL ISOPODA BY ERNEST E. UNWIN 49 and who found them homologous in structure to the branchi:e, " The ventral air-ehamber has corresponding relations. In with a simple ·blood cavity bounded by e.pithelium. The " other words, in the special part of the gill there is a space i<nner surface of this has a very thin cuticle. Kimus describ " between the general wall of the gill (composed of chitine ing the exopodite of Asellus aquaticus says:- "and hypoderm) and the boundary wall of the blood-cavity, "and this space contains air. This s;pace does not com "Rappelons que 1es parois des cavites •sanguines sont ,, municate by any opening with the outside of the gill." "formees dans Ia plus grande parti.e de leur surface, par " les deux lamelles epitheliales tapiss.ees interieurement d'une Stoller supports these observations by a number of '' mince.cuticule." experiments:- This very thin cuticle is more evident in some species, "As has already been stated, when an animal is placed e.g., Porcellio scaber, and it is this that Stoller has taken to "in water the appearance of air in the gills passes away. he a definite .cellular layer of mesoblastic origin. T·his mis " This observation led to a number of exp,eriments, the chief take runs all through his treatise, his figures too· are incor "purpose of which was to gain evidence, in addition to the rect in this particular, :but perhaps the most serious of the "evidence derived from direct inspection and from the study errors into which he has fallen has reference to the minute " of the structure of the gill, that air is normally present in anatomy of the s.pecial part of the exopo·dite; serious because " the chambers. it has found its way into such a textbook as "A Treatise on 1. "An animal was. .p laced in water until the appear " Zoology," Part VII. " ance of air in the gills. had passed away. It was then He divides the internal cavity of this part into five '' killed, and the tissues fixed by .hot 33 .per cent. alcohol. In separate chambers, one below the other, thre·e of these " sections of the gills prepared from this specimen the air are blood chambers and two are air chambers. " .chambers did not appear empty (as usual), but contained ''blood without corpuscles, that is, blood plasma (appearing He states (pp. 26 and 27) :-"The dorsal and ventral "as granulated matter). • "blood chambers of the general part of the gill are con " tinued into the special part of the gill. But in the special "This experiment, having indicated that under the con " part there is pres.ent, in addition, a. third blood-chamber, " ditions imposed the air was replaced in part, at least, by "which lies between the other two. This middle chamber is " .plasma of blood, we next sought to modify the experiment "formed by the forking of the partition separating the " in order to ascertain whether the presence of blood in the "•dorsal and ventral chambers in the general part. Each " chambers was due entirely to the conditions being "branch 'Of the fo.rk retains the structure of the main parti " artificial. " tion; that is to say, consists of a double wall, with nuclei 2. "The pr·eceding experiment was repeated, excepting " lying in the narrow intervening space. While the dorsal "that the animal was placed in normal salt solution instead "and ventral chambers are thus separated along their lateral "·of water. The result was that the chambers contained "faces they communicate with each other in certain places "blood plasma, but perhaps less in quantity than in thB "at the margin of the gill." " previous case. Then follows a description of the possible communica "In this case the two fluids were of approximately equal tions betwe·en the dorsal and ventral chambers, and this so "de.nsities and, it may be inferred, tended tCJ Teplace the air called middle chamber, whkh I ne·ed not quote. He then "in equal measure. Allowance must be made, however, for proceeds:- " the greater thinness of the wall bounding the blood-cham "In addition to the three blood-chambers in the special " ber than the outer chitinous wall. "•part of the gill there are other chambers which contain " air. These are two in number, situated one on the dorsal 3. "The ex-periment was re,peated, using a concentrated "and one on the ventral ·side of the gill. The dorsal air " salt solution. The result was that the chambers appeared " chamber is bounded outwards by the wall of the gill and "as empty or nearly so, little blood plasma having entered "inwards by the outer wall of the dorsal blood-chamber. "them. ~ 50 RESPIRATORY ORGANS OF TERRESTRIAL ISOPODA BY ERNEST E. UNWIN 51 "In this case it would appear that the more concentrated special part, but only outside. The following simple obser "fluid replaced, in the main, the air. In all three cases the vation could be made as well. Take off any one of the exopo " results became intelligible on the supposition that the dites from a living specimen, dip it into alcohol to remove "chambers contain air. Thus, these experiments warrant the surface air which clings to the corrugated surfaces, and " the inference that the chambers normally contain air and mount it in water. This need not take more than 30 seconds ,; not iblood." (Stoller.) at the most. The special part of the exopodite is quite trans~ Now let me take some of the points raised in the above parent, there is no trace of air when seen by transmitted excerpt- or reflected light. Now, if there were air-chambers, as Stoller suggests, it would have tbeen impossible for the air ! do not agree with the description of the partition in and water or air and alcohol to have interchanged so rapidly, the sp·ecial part. In most sections the partition seems to me especially as Stoller agrees there are no openings, and that to be a single structure composed of two closely applied very any interchange must take place by diffusion. thin strands containing elongated nuclei. In some sections the two strands are rather more separated than in others, Having convinced myself that Stoller is wrong in his but I cannot find anything which can be called a middle account of these air-cavities, there is no need to follow him chamber. The partition here does not differ materially from further, and point out the errors in .his account of their the partition in the ordinary part, indeed the two partitions physiological function, for it is base'li upon error. are continuous. I 'have failed absolutely to find the two Although claiming for these ·special breathing organs a air-chambers either by surface observation or by sections simpler structure, they are still very wonderful adaptations or by the ex:periments suggested by Stoller. and illustrate another method whereby gill plates can be used In the paragraph from .his treatise quoted above, he for air-breathing. It is interesting to find these adaptations describes these chambers as being bounded by the epithelium and to compare them with similar contrivances in other of the outer wall on one side and the wall of the blood cavity members of the Arthropoda. It has seemed to me that the on the other. We have already seen that he is in error in lung books of .Spiders may have much in common with the describing a definite boundary wall to the blood cavities, and branchial plates of land Isopods. this initial mistake may have led him into the further error Before leaving the question of the exopodites, I want to of finding air within the appendage. Certainly there is no point out a peculiarity associated with the last pair of plate wall between his so-called air-cavity and the blood cavity: like abdominal exopodites (that is of the 5th abt:!ominal it was not surprising, therefore, to find blood in these cavities. appendage). The under or dorsal surface of one of these I have tried every way I could devise to test the question, appendages is shown in Fig. 10. T~here is a curious arrange but have failed to find any air inside the appendage, and I ment of long spines or hairs, whi·ch occur in radiating groups feel convinced that his observations are incorrect, and that in a special 'band across the exopodite. The arrangement the internal cavity of the special part of the exopodite is suggests a cheval-de~frise, especially as it occupies the place, simply divided into dorsal and ventral blood chambers. In just posterior to the endopodite, where protection might be all my sections I have found blood corpuscles in these cham necessary. It is difficult to .be sure as to the function of this bers. Possibly the idea arose after looking at a section in formidable array of hairs. Protection from enemies or from which the proto.plasmic lining had contracted leaving a space the "excreta" discharged from the anus, which is close by, betw.een it and the chitinous cuticle. This does take place do not seem sufficient, and their absence in Trichoniscus is due, I suspect, to a too rapid method of dehydration or faulty against these uses. The only suggestion which occurs to me fixation-and this appearance would certainly result after as likely is the possible use in preventing the drying up of the use of salt solution. the film of water which covers the endopodites, seeing that Really a few minutes' observation of a living creature the last pair are more exposed when the plates are lifted and when the en'li of the abdomen is flexed. submerged in water and so arranged that the ventral sur face of the abdomen can be viewed in a good light would The Endopodite: The endopodites of Oniscus are smaller convince anyone that there is no air inside the walls of this in proportion to their exopodites in surface area than those