164 BULLETIN OF THE 



the embryonic polypite. A larva of Agalma in a stage following the Athorybia 

 stage bears a remote likeness to the genus Physophora. Although the resem- 

 blance is somewhat distant, for want of a better name I have called it the' 

 Physophora stage (Plate IX. fig. 1). 



In this stage the tasters are arranged in a circle on an enlargement of the 

 axis of the larva opposite the end which bears the float. This is a true char- 

 acteristic of Physophora. Although nectocalyces are well formed, there is a 

 section of stem between this terminal enlargement of the axis and the lowest 

 nectocalyx which bears covering scales. This last feature separates the young 

 Agalma from the genus Physophora. On the same enlargement which bears 

 the circle of tasters there are two polj^ites, an embr)'onic, which is the modified 

 yolk sac with the tentacle from which embryonic knobs are pendent, and a poly- 

 pite with the characteristic tentacle and knobs of the adult. Both of these arise 

 from the axial enlargement at the end of the stem. From a point on the axis 

 just below the lowest nectocalyx hangs a single taster with tentacle, and small 

 buds which later grow into polypites. 



Tentacular knobs of both kinds coexist in this stage, but they are never 

 found together in Agalmata in which there are more than four pairs of necto- 

 calyces. No provisional or embryonic organs appear in stages between the 

 Physophora stage and the adult Agalma. As far as the anatomy of the adult 

 Agalma is concerned, I have little to add to what has already been given by 

 others. In the arrangement of difi"erent individuals on the stem there is always 

 a definite sequence, and the different individuals are never displaced from their 

 proper order. Nectocalyces are always found on the nectostem, while feeding 

 polyps, tasters, and sexual bells follow in an order which is exactly reproduced 

 in different sections of the polyp stem. If we take a single such section the 

 order is found to be as follows. Beginning ■ndth the upper end, there is found 

 at first a polj^oite, just below which is the grapelike cluster of female bells. 

 Removed by a considerable space on the stem from these, there is a cluster of 

 tasters surrounded by male bells, and then, after another interval of about the 

 same length of stem, another polypite with female bells and the beginning of a 

 new section, which if followed out would be found an exact repetition of the 

 preceding. This sequence is normally followed, whatever the length of the 

 stem may be. New members of the polyp stem arise in the region just below 

 the lowest nectocalyces. New nectocalyces always form on the nectostem just 

 below the float. 



In the Agalma which is figured in Plate X. there are seventeen pairs of nec- 

 tocalyces, and seventeen sections bearing polypites and female sexual ])ells. 

 This numerical identity is not a coincidence, but seems to occur normally in all 

 stages of growth after that called the Physophora larva. 



The development of the adult feeding polyp or polypite of Agalma seems to 

 be quite peculiar. The feeding polyp originates as a simple two-layered bud 

 from the stem, and assumes at first a glol)ular shape. From this it elongates 

 into a flask-like body, the proximal portion of wliich retains a spherical form, as 

 shown in Plate IX. fig. 6. This spherical basal part is formed almost entirely 



