How far this synonymy is valid, we cannot decide. Bedot had the opportunity of 

 seeing many living specimens of Crystallomia near Amboina 95 but he does not give any 

 details of their structure. Mertens' figures of Agalina Okenii, which Brandt called Agahna 

 Mertensii 35, were not published, and we have had no opportunity of seeing the original 

 drawings. Judging from the shape of the bracts, such as Dana 58 describes, Gegenbaur's 

 fragments of Agalma Okcnii 60 and Fewkes' complete specimens of the same also belong 

 to Crystallomia polygonata Dana. We quite agree with Bedot, that Haeckel's new generic 

 denomination Crystallodes 69 ought to be dropped. When we compare both te.xt and fio-ures, 

 no generic differences can be noticed between the two. We hesitate to decide whether all 

 the three species of Haeckel: Crystallodes rigida, vitrca and Mertensii actually belono- to 

 Crystallo?nia polygonata. 



All the specimens of Crystallomia above-mentioned were described from beautiful complete 

 specimens, probably drawn and described after life. 



Preservation — even in formaldehyd 4°/^ — is very unfavourable for giving an exact 

 idea of the structure of Agalmidae. Our si.xty-eight specimens from the Siboga expedition are 

 all more or less incomplete ; sometimes the appendages fail, and the exterior of the specimens 

 changes especially through the contraction and torsion of the stem. In some specimens the 

 nectosome is entirely contracted, the nectophores having fallen off, and then the siphosome is 

 better preserved, and very often we find that the contrary is the case. 



We could not possibly make any drawings such as Dana 58 and Haeckel 69 and 88 b 

 give of complete examples of Crystallomia. 



Moreover our 68 specimens are much smaller, the smallest measuring iV,, mm., the 

 largest a few centimeters. A complete list of the different size of the specimens we o-ive in 

 each group apart. 



This was one of the reasons, why ever since beginning of our examination of the 

 Agalmidae, we doubted whether it would be possible to bring our smaller specimens into 

 connection with the species already known. 



We therefore do not wish to give any definite specific denomination but divide our 

 specimens into two groups, called Crystallomia spec, groujj I and Crystallomia spec, group 11. 

 They all possess at least one characteristic which also belongs to Crystallomia Dana. We are 

 convinced that even this division into two groups will not be a lasting one, but that investio-ators 

 with more complete and especially living material, will subdivide them and by accurately comparino- 

 their material with our publication will be able to introduce definite specific denominations. 

 The division into two principal groups is based solely on the shape of the tentilla. 



On superficially comparing our sketches (PI. XII, PI. XIII, PI. XIV, PI. XV) with the 

 publications of Dana 58, Gegenbaur 60, Haeckel 69 and 88b one would not at first sio-fit 

 find any difference between them. We find, however, that there are many differences. The 

 material of the Siboga expedition taught us how all the difterent tentilla such as the above- 

 mentioned authors describe, are represented in the Siboga material and we have even to describe 

 two other new different kinds of these appendages which occur in different specimens and durino- 

 special periods of development. 



