APPEARANCES DECEPTIVE

137

Family 2.-Homolido.

The carapace is quadrangular or subtriangular. The eye-stalks are usually slender and very long, the orbits very incomplete. The first antennæ are not retractile into special fossettes. The last pair of legs are small, prehensile, subdorsal in position.

The family may include five genera. The species extend to moderate depths.

Dicranodromia, A. Milne-Edwards, 1880, is what is called an inosculant genus. By the character of the last two pairs of legs it should belong to the Dromida, but the defective orbits and the want of fossettes for the first antennæ place it among the Homolidæ.

Homola, Leach, 1815, has the carapace quadrilateral, of greater length than breadth. The eye-stalks have a long slender basal part and a shorter dilated corneal portion. The chelipeds are of moderate size, the three following pairs of limbs are long and flattened, while the last pair are short and subchelate. Homola barbata (Herbst) and Homola Cuvieri, Risso, occur in the Mediterranean and Atlantic. Homola orientalis, Henderson, is a Pacific species.

Latreillia, Roux, 1828, has a triangular carapace. The eye-stalks are very long and slender, cylindrical, turned forwards, and divergent. The legs are slender and cylindrical, the three middle pairs being very long. The four species belonging to this genus are apportioned two to Japan, one to Australia, and one, Latreillia elegans, Roux, to the Mediterranean and Atlantic. The figures in Plate V. represent the Japanese Latreillia valida of de Haan, from whose work they are copied on a reduced scale. At the first glance any one would be tempted to place the genus among the spider-crabs in the tribe Oxyrrhyncha, but de Haan showed the impropriety of this. The same structure,' he says, 'which prevails in Dromia and Homola is found in the species of Latreillia. They agree in the organs

of feeding, motion, and generation, and in the fabric of the trunk within. The vulvæ are not in the trunk, but in the bases of the third pair of legs. The sternum is deeply excised between the last pair of legs; the turkish saddle1 and median apodeme are wanting; the transverse septum of the apodemes forms in the middle a central yoke and central canal. Further, they show a peculiar affinity to Homola by the eye-stalks, which are very long, thin, apically inflated, as good as free,2 by the first antennæ not retractile into fossettes, by the first and third maxillipeds, which so agree in the two genera that they can scarcely be distinguished. With the Maiacea, therefore, Latreillia cannot be confounded, although the form of the trunk is triangular, the legs are very slender, the epistome is quadrate, and of the branchiæ there are ten pairs, of which three are connected with the maxillipeds, two with the chelipeds, and three with the following legs, but the last united with the penultimate legs.'

Latreillopsis, Henderson, 1888, with a type species, Latreillopsis bispinosa, from the Philippine Islands, ' occupies an intermediate position between the genera Homola and Latreillia. From Homola it is distinguished by the arrangement of the rostrum and supraorbital spines, the greater length of the ocular peduncles, and more especially by the elongated cylindrical legs. In Latreillia, on the other hand, the frontal region is narrow and produced so as to give the carapace a triangular outline, the supraorbital spines are more strongly developed, and the eyestalks and legs are of greater length.' It may perhaps be regarded as something more than a mere coincidence that this link between Homola and Latreillia was obtained by the Challenger in one of the two localities in which the same vessel took specimens of those two genera.

Homologenus, A. Milne-Edwards, 1888, bears a name altered from the pre-occupied Homolopsis, 1880. It differs

'The expression posterior turkish saddle,' is applied by MilneEdwards to the small arch formed by the sternal apodemes which spring from the hind margin of the last segment of the trunk.

2 The Latin is totis quantis liberis.

from Homola in having a more ovoid carapace, a more developed rostrum, feebler legs, and especially in the form of the eyes which are very small and not narrowed at the base.

Here also should perhaps be placed some Australian species of the genus Paratymolus, Miers, 1879, in which the carapace is deflexed in front, flat behind, with the sides nearly straight, the 'front' being prominent and narrow. For a new family Paraty molida Mr. Haswell gives the characters, 'carapace in general form similar to the Maioidea. External maxillipedes partly over the epistome.' He thinks that it would perhaps be better placed among the Corystoidea. Mr. Miers, however, does not agree with this view, but thinks that it ought to stand near the Dromidæ. It is an argument the more for including the present tribe in the Brachyura, that two experts should be unable to agree whether a genus belongs to it or another tribe which is brachyuran beyond question.

Legion 2.-Ranininea.

The carapace is ovate-oblong, with the regions not defined, and the 'front' of varying width. The orbits are well marked. The first antennæ are without special fossettes, and are placed to some extent behind the second pair. The third maxillipeds are moderately elongate. The sternal plastron or breastplate is wide anteriorly. The walking-legs have the terminal joint broad and compressed; the last pair of legs are small and subdorsal in position. The vasa deferentia of the male are protruded. The pleon is short, partially extended, not folded under the trunk, with four pairs of appendages in the female.

Dr. Henderson includes the epithet 'smooth' in the description of the carapace, but this is obviously unsuited to Ranina scabra.

Dr. Boas has ingeniously suggested that the position of the vulvæ in the bases of the legs instead of in the sternal plastron has been brought about by the extreme narrowing of the plastron, and this may well have been

so, though such a species as Petalomera pulchra in the preceding legion does not seem to suit the theory. One may, however, suppose that in some instances a re-widening of the plastron may have been developed without any rearrangement as to the position of the vulvæ.

This legion contains the single family Raninidæ, for which, therefore, no separate character is needed. It includes some nine genera, limited to the warm seas, and inhabiting chiefly the tropics, with a range of depth apparently. not exceeding 300 fathoms.

Ranina scabra (Fabricius), originally called Cancer raninus, Linn., and afterwards Ranina serrata, Lamarck, and Ranina dentata, Latreille, from Amboina and the Sandwich Islands, was known to fame long before a separate genus was established for it. The carapace has been compared to an inverted triangle. It is very broad anteriorly, but the sides slope very gradually to the rounded hinder margin. The eye-stalks are three-jointed, strongly geniculate, and have a very deep orbit. The pterygostomian regions of the carapace unite with the sternal plastron so as completely to separate the third maxillipeds from the chelipeds. The plastron itself is anteriorly almost trefoil-shaped, but to the rear becomes linear. The branchia, Milne-Edwards says, are arranged as in the Brachyura, but in the conformation of the respiratory cavity there is a peculiarity of which he knew no other example. As in the Leucosiidæ, the carapace is joined to the sternum and the cavity of the sides, without leaving above the base of the feet or maxillipeds any space for the entrance of the water necessary for breathing, but the afferent channel instead of being pierced beside the efferent channel, on the sides of the mouth, is situated behind and has a special opening below the base of the pleon. This view, however, is criticised by de Haan in a passage that is not free from perplexity. 'In Portunus and Grapsus,' he says, 'the water is brought to the branchia by a double path and removed by a double path; it reaches the branchial cavity by the mouth and the apertures near the base of the chelipeds; but it passes out both by the space between the inferior lateral margin

A DISCUSSION ON BREATHING

141

of the trunk and the epimera and by two ducts under the insertion of the pleon. In Ethusa the case is the same.

FIG. 11-Ranina scabra (Fabricius), a male specimen, with separate figure of the pleon

In Calappa and Matuta the hinder ducts are not below but close to the pleon over the first joints of the fifth pair of