68 



NA TURE 



[May 15, 1884 



In Brooks' theory the main point is the migration of 

 ova from the solitary individual into the individuals of the 

 chain. In the light of a study of closely allied genera 

 we find serious objections to this view. The fact on which 

 it is chiefly based is that in the stolon when quite immature, 

 we can trace the following organs : a, the outer tunic or 

 epidermis ; b, the pharyngeal cavity continuous with the 

 pericardium ; and c, two " club-shaped masses" of cells which 

 lie on either side of b, and which soon resolve themselves into 

 two lines of ova, one of which passes into the sinus system of 

 each zooid. The discovery of undoubted ova in the stolon when 

 the organs of the zooids are hardly indicated suggests, says 

 Brooks, a migration from the nurse, which is therefore female. 



Now we have, in direct opposition to this, an observation of 

 Salensky's, that in some cases a second ovary is developed in the 

 chain- Salpa? ; and an indirect negation is entailed in the facts of 

 gemmation in Pyrosoma, which is generally allowed to be a less 

 modified form. Here we find the bud, when merely a protu- 

 berance on the mother, consisting of an epidermis derived from 

 that of the mother, an " archenteric" cavity continuous with the 

 endostyle, and a mass of cells which are derived directly from the 

 " generative blastema." In this mass a single ovum can be seen 

 quite as early relatively as in Salpa, and a second near the base 

 supplies the secondary bud. Among the Composite Ascidians 

 the case is similar ; in Amourcecium the buds are cut off from 

 the post-abdomen and consist of outer tunic, mesentery (that is, 

 continuation of pharynx backwards), and two lateral masses in 

 which germinal vesicles shortly appear ; in Didemnium, also, 

 although Kowalevsky traced the buds back to a condition much 

 more nearly resembling the segmenting ovum, still even here the 

 single ovum is one of the most conspicuous of the primitive 

 organs. It is apparent that there is in the Tunicata a tendency 

 to form buds at the expense of the three primitive layers, and 

 that some advantage attends the early development of ova. 

 Whether this is to avoid the danger of self fertilisation or not, it 

 reaches its limit in Salpa, where the rudiment of the ovary only 

 consists of one fully developed ovum. 



A similar modification in the time of development of the ova has 

 taken place in some of the Hydrozoa (Hydrella, Sertularia, &c.), 

 where, as Weismann has pointed out (Abstract and Review by 

 Prof. Moseley in Nature, vol. xxix. p 114), the immature ova 

 may be detected in the coenosarc before even a rudiment of the 

 bud appears. 



The view which Todaro upholds seems also to be negatived 

 by Salensky's observations. For if the solitary Salpa is deve- 

 loped by follicular budding, it is not remarkable that some of the 

 cells should form an organ corresponding to the generative blas- 

 tema of Pyrosoma, i.e. giving rise only to the generative 

 cells. 



If Salensky's facts stand the test of further observation, we 

 have in Salpa not only a unique method of development but a 

 unique alternation of generations, namely, of gemmation and 

 parthenogenesis, only comparable to that of the Aecidiomycetes 

 among plants. 



In Pyrosoma, the individual developed directly from the egg is 

 the " cyathozooid," and this remains rudimentary, giving rise to 

 the first four ordinary individuals by budding. There is here an 

 alternation of a single asexual form with numerous generations 

 produced by budding, each of which becomes hermaphrodite. 



According to Ganin, we have a precisely similar case in the 

 Composite Ascidians, for he states that sexual organs only develop 

 in the individuals produced by gemmation. 



In Doliolum the ovum, as before, gives rise to an asexual indi- 

 vidual, but the lateral and median buds which arise on its dorsal 

 stolon do not become sexually mature. The former only serve 

 to nourish and protect the latter (Grobben, Arbeit, der Zool. /"- 

 sti'. SB Wien, iv. ), from which a ventral stolon (stalk of attach- 

 ment) grows out, bearing sexual individuals. 



Outside the group of the Tunicata, true alternation of genera- 

 tion occurs in some Ccelenterates, some Annelids, some Cestodes, 

 and possibly in the Trematodes. 



The alternation of hydroid and medusoid forms in many 

 Hydromedusae (Gymnoblastic and Calyptoblastic Hydroids and 

 Hydrocoralla), all Acraspeda except Pelagia, and possibly Fungia 

 among the Actinozoa, has been dwelt on in a previous paper in 

 this journal by Prof. Moseley, and little range of modification 

 occurs except in the extent of development of the medusae. 



In Cestodes the complicated metamorphosis has been mistaken 

 for metagenesis, but it is only in those cases where the cystic 

 worm develops a number of "heads" by gemmation, e.g. 



Ccenurus cerebralis. Tun:. us, that metagenesis really 



There is more uncertainty about the condition of affairs in the 

 Trematodes, the ordinary view being that "the majority of the 

 stages are simply parts of a complicated metamorphosis, but in 

 the coexistence of larval budding (giving rise to Cercariae or a 

 second generation of Rediae) with true sexual reproduction there 

 is in addition a true alternation of generations " (Balfour, " Com- 

 parative Embryology," vol. i. pp. 172, 173). Grobben (loc. cit.), 

 however, has lately stated that the Cercariae are developed not 

 from a mass of cells produced by internal budding in the Redia 

 but from an ovum developed parthenogenetically. This would 

 place these phenomena under the category of heterogamy. 



Among the Polychaetes there exist in Syllis, Myrianida, 

 and Autolytus undoubted cases of alternation of generations ; 

 but these are not of recent discovery, having been described by 

 Qnatrefages, Krohn, and A. Agassiz about 1850-1860 (Balfour, 

 ibid., i. pp. 2S3, 284). 



A general comparison of the various ways in which reproduc- 

 tion is carried on within the limits of alternation of generations 

 may be easily made by a series of diagrams in which A represents 

 the asexual individual developed from the fertilised egg, B the 

 sexual zooid, and A B those forms which carry on both methods 

 of reproduction. 



Pyrosoma and 

 idia 

 A cyathozooid 



Salpa 



A solitary ? partheno- 

 genelic 



\ \ 



. Hydrozoa 



A A 1 hydroid 



Doliolum 



A with dorsal stolon 



|\ 



\L'lateral 

 M median bud 



B sexual zooid 



Fungia and Acraspeda 

 A scyphistoma 



/ I \ 



B II B ephyroid 



Heterogamy, which is not so common as metagenesis, has been 

 the subject of very interesting memoirs by Adler and Lichten- 

 stein. In a paper, " Ueber den Generationswechsel der Eichent 

 Gallwespen " (Jenaische Zeitschrift, 1881), we have the result of 

 Dr. Adler's work on "Gall-making Hymenopterous Insects," 

 formerly described as belonging to eight different genera, namely 

 Neuroterus, Aphilotkrix, Dryophanta, Biorhha, Spathegaster, 

 Andricus, Teras, and Trigonaspis. He confirms the conclusions 

 of Lichtenstein that certain species of the first four of these 

 genera are stages in the life-history of certain species of the last 

 four. 



The gall wasps which in April leave the small round scale-like 

 galls on the under surface of the leaves of the oak, have been de- 

 scribed as Neuroterus ventricularis ; but the parthenogenetic egg 

 develops within a round soft "currant-gall" to a wasp named 

 Spathegaster baccarum. The latter escapes in June, and differs 

 from the Neuroterus in size, colour of the legs, and in the female 

 in the number of joints in the antennae. The eggs produced by 

 the Spathegaster when fertilised develop within Neuroterus galls. 

 A still greater difference exists between the two generations 

 formerly called Biorhiza renum and Trigonaspis crustalis. 

 Trigonaspis is 4 mm. long, winged, almost entirely black, with 

 antenna; of 15 ( £ ) and 14 ( 9 ) joints, while the Biorhiza is 

 I'S mm. long, wingless, red-brown, and with 13 joints in the 

 antennae ; the two forms live, moreover, in different kinds of 

 galls. In all these cases the alternation is direct. But among 



