second maxillipeds, where it is held in place by the flat oval chela which are tightly clasped over it. At the opening it stretches out as far as it can reach without leaving the burrow, and dropping the armful of sand it smoothes it down until it is level with the surrounding surface. This process is then repeated until the burrow reaches a great depth, for I have dug for three or four feet without reaching the end, and all the specimens which I kept in confinement burrowed to the bottom of the aquarium. When the burrow is finished the animal spends most of its time near the top, and as the semicircular exopodites of the abdominal appendages complete the outline formed by the convex dorsal surface, it completely fills the circular tube, into which the constantly vibrating scoop-like abdominal appendages carry a continuous current of water, which escapes through the loose

sand.'

Pseudosquilla, Dana, 1852, has a name originally coined but not published by Guérin.

In this genus the terminal joint of the second maxillipeds is without basal enlargement and with few marginal spines or none; the pleon is smooth, convex, and narrow; the terminal joint of the first pleopods of the male imperfectly divided by a marginal notch into an inner and outer lobe; the telson has the submedian spines long and tipped with movable spinules, and usually it has a single secondary spinule, but sometimes two, three, or four such spinules between the submedians and intermediates. Pseudosquilla ciliata, Miers, a species found alike at Honolulu and in the West Indies, is of a bright cherry red colour.

Gonodactylus, Latreille, 1825, has the terminal joint of the second maxillipeds enlarged at the base, and without marginal spines; the pleon narrow, convex, thick; the primary marginal spines of the telson very large, with one or two secondary spines between the submedians and intermediates.

Gonodactylus falcatus (Forskål), according to Kossmann, must supersede the better known name Gonodac

NAMES GIVEN WITH A PURPOSE

287

tylus chiragra (Fabricius). The species appears to have a vast range, and its colour is said to be exceedingly variable.

Coronida, Brooks, 1886, has the carapace flat and nearly rectangular. The rostrum ends in a small median spine; the scale of the second antennæ is very small; the terminal joint of the second maxillipeds is dilated at the base, and armed with spines on the inuer margin. The pleon is depressed; its hind segments and the telson are thickly set with small spines. The uropods are very small. The name of the genus is compounded from the names of the rejected genera Coronis and Chlorida (or Chloridella), to indicate Professor Brooks's view that like them it contains somewhat primitive species. These are Coronida Bradyi (A. Milne-Edwards) and Coronida trachūrus (Miers), both of which were originally assigned to Gonodactylus. The professor seems inclined to suspect that the sixth pleon-segment may prove to be fused with the telson, but the type-specimens of Miers's species in the British Museum show that at least in Coronida trachurus the telson is quite distinct from the preceding segment and freely movable upon it.

Protosquilla, Brooks, 1886, has the rostrum furnished with long acute median and anterolateral spines; the eyes and the scale of the second antennæ small; the terminal joint of the second maxillipeds dilated at the base, without marginal spines; the pleon convex, its sixth segment more or less completely fused with the telson; the uropods small. Professor Brooks considers the name of this genus 'the more appropriate inasmuch as all the other Stomatopoda present evidences of divergent descent from a common stem form, which, like the living representatives of the genus Protosquilla, was characterised by the small size of its eyes, antennary scales, and uropods.' Seven species are included in the genus, among which Protosquilla elongata, Brooks, with a carinate and bilobed but otherwise simply constructed telson, presents a rather striking contrast to Protosquilla Guerinii (White), in which the dorsal surface of the telson carries twenty-two

long spines symmetrically arranged, while the preceding segment, which in the other species has some rather inconspicuous carina, here has about fifty-six long cylindrical spines, each of which ends in a blunt rounded tip with a perforation from which a soft tubular fleshy process protrudes.'

Leptosquilla, Miers, 1880, is defined as having the 'ophthalmic segment greatly elongated; rostrum not reaching beyond half the length of this segment,' and to it is assigned the single species Leptosquilla Schmeltzii (A. Milne-Edwards), from the Samoa Islands.

For a history of the development of the Stomatopoda the foundation was laid by Dr. Claus in 1871. Many additions have been made and much precision given to the knowledge of this subject by the researches of Professor Brooks. The transparent pelagic larvæ are frequently captured, and in consequence of their glass-like clearness the individual organisms can be easily studied, but there are special difficulties connected with tracing the series of changes which they pass through between the egg and the adult condition. Unlike most Malacostraca, the Squillida do not carry their eggs about with them, but deposit them in their deep burrows, there to be aerated by the currents of water which the pleopods of the parent maintain. Hence the hatching of eggs in an aquarium has not yet been brought about. The older larvæ are hardy, but seldom found near the shore. The younger larvæ can be found near the shore, but seldom moult in confinement. The life-history has, therefore, to be traced by a comparison of numerous forms collected from the open sea, and here there is a twofold risk, of uniting the larvæ of quite distinct animals as stages in the life-history of a single species, and of disuniting the larval stages of a single species as though they belonged to distinct species or even distinct genera. In this labyrinth the explorer has to be guided in part by general resemblances, but more especially by comparative measurements. Professor Brooks, having obtained a number of specimens which by general resemblance seemed to make

SCIENCE IS MEASUREMENT

289 a good series, proceeded to tabulate the measurements in millimetres. He was disappointed, however, not to find any such conformity to a general law as he had expected. It then occurred to him that in the larva change of form and increase of size are going on together, and that one organ might be diminishing relatively to another organ while it was increasing relatively to itself in a former stage. He therefore reduced his measurements to a common standard, expressing them in thousandth parts of the total length of the larva at each stage, and this usually enabled him to decide whether a given larva did or did. not belong to a particular series. Apart from this, in one instance which he mentions his comparative measurements obtained a triumph parallel in its way to that of Kepler when he found that the positions of the planets conformed to a numerical law. Having a series of larvæ apparently of the so-called Coronis (Erichthus) minutus, of which the youngest measured 4.16 millimetres, the second 5.29 mm., the third 6.49 mm., and the fourth 10.21 mm., he observed that if 4-16 is multiplied by five-fourths, and the result by five-fourths, and so on, the resulting series of numbers is

(1) 4·16, (2) 5·20, (3) 6·50, (4) 8·13, (5) 10∙16,

which corresponds as exactly as need be with the series of larvæ, allowing for the absence of the fourth stage. It is highly improbable that larvæ bearing a general resemblance to one another would have this curious numerical relationship unless they also had a very near relationship by blood likewise. Professor Brooks considers that 'the free prolonged larval life has brought about modifications which have no reference to the life of the adult, so that the larvæ differ among themselves more than the adults do.' By reason of their small size and great transparency some of these larval forms are very suitable and interesting objects for the microscope.

The following list will show the names attached to the young forms of the different genera:

In support of this affiliation Brooks says, 'Faxon has reared a Squilla empusa from an Alima larva. The Challenger collections show that the larva of Lysiosquilla maculata is one of the short-spined Lysioerichthi; I have reared Lysiosquilla (Coronis) excavatrix, one of the more primitive Lysiosquillæ, from a long-spined Lysioerichthus. Claus has figured a series which shows beyond question that at least one species of Pseuderichthus becomes a Pseudosquilla, and the Challenger collections furnish equally good proof that the Chiragra group of Gonodactyli come from Gonerichthus larvæ, and I therefore believe that we may very safely assume that all the Lysioerichthus larvæ are young Lysiosquillæ, all the Alima larvæ young Squillæ, Alimerichthus one of the lower Squilla or Chloridella, all the Gonerichthi young Gonodactyli, Erichthalima very probably a young Coronida, and all the Pseuderichthi very probably young Pseudosquillæ.