maximum time. Thus for the circuit under consideration, and with successive repetitions of the current in the same direction, it takes longer time for the current produced by an impressed E.M.F. of 4 volts to reach 95 per cent. of its maximum than it takes for the current produced by either 3 or 5 volts to reach 95 per cent. of their maximum. The results show also that, within considerable limits, the time required for the current to become uniform is on the whole nearly inversely proportional to the impressed E.M.F., and that for moderate values of the E.M.F. the time may be very great; when the E.M.F. was 2 volts, and the current sent in such a direction as to reverse the magnetism left in the magnet by a previous current of the same strength, the time required for the current to establish itself was over three minutes. The difference of time required for repetition and for reversal of previous magnetisation was also very marked when the iron circuit was closed. The results show that great errors may arise by the use of ballistic methods of experiment, especially when weak currents are used, and that for testing resistances of circuits containing electromagnets, a saving of time may be obtained by using a battery of considerable E.M.F. Another set of experiments gives the effect of successive reversals of the impressed E.M.F. at sufficient intervals apart to allow the magnetisation to be established in each direction before reversal began. In this set also the effect of cutting out the battery and leaving the magnet circuit closed is illustrated, showing that several minutes may be required for the magnet to lose its magnetism by dissipation of energy in the magnetising coil. The effect on these cycles of leaving an air space in the iron circuit is also illustrated. It is shown that a comparatively small air space nearly eliminates the residual magnetism and diminishes considerably the rate of variation of the coefficient of induction and the dissipation of energy in the magnet. Several cycles are shown for the magnet used as a transformer with different loads on the secondary. The results give evidence that there is less energy dissipated in the iron the greater the load on the secondary of the transformer. Some experiments are also quoted which go to show that the dissipation of energy due to magnetic retentiveness (magnetic hysteresis) is simply proportional to the total induction produced when the measurements are made by kinetic methods. Reference is made to the recent experiments of Alexander Siemens and others which seem to confirm this view. IV. "On the Development of the Stigmata in Ascidians." By WALTER GARSTANG, M.A., Jesus College, Oxford; Berkeley Fellow of the Owens College, Manchester. Communicated by A. MILNES MARSHALL, M.A., M.D., F.R.S. Received April 14, 1892. The respiratory organ or pharynx of the Tunicata exhibits a great amount of variability in form, size, and complexity of structure in the different members of the group-a variability which is obviously correlated with the physiological value of the organ. In the simplest forms, the Perennichordata or Appendicularians, the pharynx is a mere hollow cylinder, provided with a single pair of tubular gillclefts, one on each side. In the higher forms (the Caducichordata) the cavities of the two gill-clefts become enormously dilated, so as to constitute a pair of peribranchial chambers interposed on either side between the pharynx and the body-walls. The dilatation of the gillclefts to form peribranchial chambers can be aptly compared with the formation of branchial pouches in the tubular gill-clefts of Marsipobranch Fishes. But whereas the respiratory surface of the branchial pouches of Marsipobranchs is increased by folds of the walls of the pouches, the same purpose in the higher Tunicata is effected by a different means. The inner (visceral) walls of the peribranchial chambers apply themselves closely to the wall of the pharynx, and perforations then appear at numerous points, where the pharyngeal and peribranchial membranes have actually united. The remnants of the pseudo-coelomic cavity, enclosed between the pharynx and the visceral walls of the peribranchial chambers, become extensive channels for a vigorous circulation of the hæmal fluid. The organ formed by the union of the pharyngeal and peribranchial walls is usually referred to as the "branchial sac;" and the perforations, which put the cavity of the pharynx into extensive communication with the peribranchial chambers, constitute the so-called "stigmata." The stigmata vary much in form and arrangement. In the fixed Ascidians, whether simple or compound, they are usually simple slits, of a narrow elongated form, arranged in a series of rows placed transversely to the longitudinal axis of the body (Ascidia, Clavelina, Botryllus, Styela). In some genera, however (Corella, Molgula), the stigmata are curved and somewhat spirally arranged; but this condition is undoubtedly derived from the former by modifications of a secondary nature. In the pelagic Tunicata (Salpa, Doliolum, Anchinia, Pyrosoma), the condition met with in the fixed Ascidians is never found; there is never more than one row of stigmata on each side, and this row, though occasionally oblique or even transverse, is usually longitudinal in direction. In Pyrosoma and some species of Doliolum the stigmata are narrow, elongated slits, extending transversely across the whole lateral face of the pharynx-each stigma occupying an area which in the fixed Ascidians is taken up by an entire transverse row of stigmata. To ascertain what is the fundamental order underlying all these variations, and to determine what degree of correspondence and homology there is between the stigmata of the pelagic Tunicata and those of the fixed Ascidians, is not an easy matter; indeed, the possibility of any detailed comparison hardly seems to have occurred to the majority of investigators. Two views, due to Herdman and Lahille respectively, are, however, worthy of mention here. Professor Herdman* derives Pyrosoma from a group of the compound Ascidians, through the curious colonial form Colocormus Huxleyi; and, in harmony with this view, he regards each of the transverse stigmata of Pyrosoma as corresponding to an entire transverse row of stigmata in the Ascidians, the several stigmata of the row having apparently coalesced to form the single stigma of Pyrosoma -a process which has almost certainly occurred in certain deep-sea types (Fungulus, Culeolus, &c.). Lahillet characterises the latter portion of Herdman's view as a "profound error," and attempts, instead, to establish the remarkable proposition that the longitudinal row of transverse stigmata in Pyrosoma is strictly homologous with one of the transverse rows of longitudinal stigmata in an Ascidian, through a phylogenetic rotation of position. The oblique position of the row of stigmata in some species of Doliolum (e.g., D. Ehrenbergi) is regarded as an intermediate condition between the two extremes. Lahille bases this proposition upon the changes of position which the organs of a Pyrosoma-bud undergo during development. These changes, it is true, are very remarkable, but they furnish absolutely no evidence for Lahille's contention; for it is a well-established fact that immediately after their first appearance the stigmata of Pyrosoma begin to elongate in a direction at right angles to the long axis of the endostyle, and this relation is maintained through all the curious changes of form which the bud undergoes in its further development. Lahille's homologies are consequently without foundation; and, although Herdman's comparison is far more justifiable, yet the development of the transverse 'Challenger" Reports," "Tunicata," 2nd Report, pp. 319, 320; 3rd Report, pp. 20, 24, 25, 137. Recherches sur les Tuniciers,' Toulouse, 1890, pp. 59, 61, figs. 44-51. Seeliger: "Zur Entwickelungsgeschichte der Pyrosomen;" Jenaische Zeitschrift,' vol. 23, 1889, p. 622, figs. 15, 17, 19. Salensky: "Beitr. z. Entwick. d. Pyrosomen;" Zoolog. Jahrbüch., Abth. f. Anat.,' vol. 5, 1891, p. 32. Salensky's figures 2 and 3, on p. 9, are, however, strangely inaccurate in this respect. stigmata in Pyrosoma clearly tends to show that these are simple structures, and that they have not arisen by the modification of so many transverse rows of stigmata, as his theory demands. The question is thus seen to be still unsolved, and if in this communication I venture to offer some observations upon the matter, it is in the belief that they tend considerably to elucidate the problem. The view that the pelagic Caducichordate Tunicata have been profoundly modified in structure through their mode of life, and that they are to be derived phylogenetically from the so-called Compound Ascidians, is at present held, with varying reservations, by almost every recent investigator of the Tunicata, except Seeliger. It is held by Grobben and Uljanin for Pyrosoma, Salpa, and Doliolum, by Herdman for Pyrosoma, by Lahille for Pyrosoma and Doliolum, and by Salensky for Pyrosoma and Salpa. The evidence for this view has always seemed to me to be very slender and unimportant; and I believe it is this widely-spread conception which is answerable, among other things, for the absence of any satisfactory comparison between the stigmata of the fixed Ascidians and of their pelagic allies. I have accordingly approached the question from the reverse point of view, believing that, by a study of the development of the stigmata in the fixed Ascidians, recapitulative stages would be met with which would furnish the desired answer. A grant awarded me by the Government Grant Committee last year enabled me, during the summer, to collect suitable material at the Plymouth station, and my observations have been completed in Professor Milnes Marshall's laboratories at the Owens College. The development of the stigmata in the larva and oozooid of Ascidians has hitherto been very little investigated. The earliest complete account is that of Krohn,* an abstract of whose observations upon Phallusia mammillata is given by Balfour (Comp. Emb.,' vol. 2, p. 20). Krohn's interpretations were subsequently criticised by E. van Beneden and Julin in their valuable papert on the "Postembryonic Development of Phallusia (Ascidiella?) scabroïdes." These investigators showed that in the latter species two stigmata at first appear, one behind the other, on each side of the pharynx, and that subsequently new stigmata arise between and behind the two first, until a longitudinal row of six stigmata is formed on each side. These stigmata enlarge transversely to the long axis of the body, and subsequently subdivide, in the order of their formation, so as to constitute a corresponding number of transverse rows of smaller stigmata. Van Beneden and Julin thus drew a distinction between primary stigmata and secondary (definitive) stigmata, and called attention to the irregular order of formation of the primary stigmata as a point * "Ueber die Entwicklung d. Ascidien," ' Müller's Archiv,' 1852. + Arch. de Biologie,' vol. 5, 1884, p. 611. VOL. LI. 2 M worthy of notice. They did not, however, draw any general conclusions from the phenomena which they observed, beyond pointing out that the irregular order in which the primary stigmata appeared was in opposition to any theory as to their metameric arrangement. The only other observations of importance are those of Seeliger* on the development of the stigmata in Clavelina. In this form, as in Phallusia and Molgula,† two pairs of stigmata at first arise, one behind the other, near the dorsal border of the sides of the pharynx. But instead of elongating in a transverse direction, as is the case in Phallusia, these stigmata elongate in a longitudinal direction, and become directly converted into the stigmata of the adult. With the downward extension of the peribranchial chambers, new stigmata arise independently, below the two first formed, so that eventually two transverse rows of perforations are formed on each side of the pharynx; and all these, by growth in a longitudinal direction, become directly converted into the slit-like stigmata of the adult. Subsequently, after the attachment of the larva, new transverse rows of stigmata arise in front of and behind the two first rows in an identical manner. I have myself followed out the development of the stigmata in Clavelina, and have nothing to add to, or alter in, Seeliger's description; the stigmata invariably arise quite independently, and I have seen no indication of such a process of subdivision as has been described above for Phallusia (Ascidiella ?) scabroïdes. A similar independent mode of origin of the stigmata has also been observed by Giard‡ in Perophora, and by Lahilles in Distaplia magni larva. Thus, up to the present time, we are acquainted with three distinct genera in which the stigmata arise independently of one another; while the process of subdivision, described for Phallusia (Ascidiella?) scabroïdes, remains unconfirmed and entirely without parallel. It would even be excusable to regard this latter method, from its exceptional character, as a developmental modification of the former. But before discussing this diversity of development, I will describe certain observations which I have made as to the development of the stigmata in several other types of Ascidians. In Botryllus the stigmata of the adult have the usual form of • “Zur Entwicklungsgesch. d. Socialen Ascidien," 'Jen. Zeit.,' vol. 18, 1885, pp. 45-150, Plates 1 to 8. + P. J. van Beneden (M. ampulloïdes). Kupffer, Arch. f. Mikr. Anat.,' vol. 8, 1872, Taf. 17, fig. 8a. Lacaze-Duthiers, 'Arch. de Zool. Exp.,' vol. 3, 1874, pp. 623, 631, Plate 27. 'Arch. de Zool. Exp.,' vol. 1, 1872, p. 677, Plate 24, fig. 6. § 'Recherches sur les Tuniciers,' p. 165. |