INSTITUTE OF SCIENCE, PHILADELPHIA. 67 



from which they sprang. A variation sometimes observed is one in wiiich the 

 cone becomes flat above, resembhng a plmiorboid coil. This is occasional in 

 Volutilithcs as an individual " sport," but is habitual with some later species. 

 Another form, developed from the trochoid, differs less in size of its initiatory 

 whorls, but contains more whorls, which increase but little in diameter trans- 

 verse to the axis. This may be termed 2. ptipiform nucleus (See PI. 6, fig. 3 a). 

 Still another important step is taken when the larval shell becomes larger, 

 bulbous or subspherical, one-sided and inflated. The number of whorls is less 

 and the nucleus is generally slightly larger than the next succeeding whorl. I 

 shall refer to this as the bulbous mccleiis. It is usually unpolished when per- 

 fect and attracts attention by its want of symmetry with the regular coils 

 of the spire succeeding it. 



Lastly we may mention the enormous nucleus observable in many recent 

 and a very few late fossil forms where the trochoid type has become greatly 

 exaggerated and the visible surface of the nucleus in the adult is dome-like 

 and as large as the end of a pigeon's egg. This differs from the trochoid 

 type chiefly in size and may be called the Melo imcleiis, as (though not con- 

 fined to that group) it is typical of the subgenus Melo as contrasted with 

 Cymba. There are all gradations of size between the Melo form and the 

 ordinary trochoid nucleus. The medium sizes are apt to exhibit transverse 

 waves or riblets and rarely faint spiral lines. 



Of course, all these names refer merely to the appearance of the outer sur- 

 face of the posterior aspect of the spire of the larva. The anterior part of 

 the shell is always canaliculate. 



Turning to the second type which begins with a horny protoconch, it may 

 be observed that the discovery of that type is very recent, and until announced* 

 the key to a full comprehension of the meaning of the nuclear shell in l^ohi- 

 tidcB was wanting. 



The origination of this group must have taken place about the end of 

 Mesozoic time. I have seen a single Cretaceous species which probably 

 should be referred to it, though the nucleus was imperfect. The group was 

 partly recognized by Conrad in 1866, but he did not at that time have data 

 for placing its validity beyond a doubt. 



In the typical form the larva is clothed with a cuticular or horny proto- 

 conch, probably similar in form to that which when shelly results in the " bul- 

 bous nucleus " of the other line of descent. Later on, but while still in the 

 ovicapsule, the deposition of limy matter begins as a slender cone or elevated 

 point along the line of the axis of the protoconch, and as the larva grows the 

 posterior part of the mantle secretes a shelly dome. Being thus cut off from 

 the horny protoconch, the latter falls into shreds and is lost. The nucleus of 

 the larva, still in the ovicapsule, then presents a slightly irregular dome, with 

 a slender point rising from the apica! part (PI. 6, fig. 5). So much is a matter 



* Dall, Blake Gastropoda, Ball. Mus. Comp. Zool. xvili. p. 452, 18S9 ; Proo. D". S. Nat. Mus. xii. p. 311, pi. ix. 

 figs. 5, 6, 1S90. 



