NO. lO AMERICAN SPECIES OF LEPIDOCYCLINA VAUGHAN J 



kind above indicated must be borne in mind ; otherwise grave errors 

 may be committed. 



The tests of orbitoids are composed of many elements, every one 

 of which is subject to variation. Unfortunately, it is not always pos- 

 sible to study the range of variation of a species, but sometimes new 

 species are, in my opinion, justifiably based on a single specimen or on 

 only a few specimens. The evaluation of the characters presented by 

 meager material is dependent on the judgment, which is a function of 

 the experience, of the investigator. For many species, however, study 

 of variation is possible. 



Some variations seem to be along definite lines and to be parallel 

 in different species. In L. waylandvaughani, L. parvula, L. pancatmlis, 

 and L. yunmgunensis, the papillae may be scattered but somewhat 

 larger over the center of the test, or they may tend to fuse or increase 

 in size with the production of knobs of variable arrangement, ulti- 

 mately grading into specimens with large costae on the umbonal part 

 of the test. For the strongly costate variety of L. parvula the varietal 

 name crassicosta is proposed, and a similar variety of L. yurnagunensis 

 is named var. morganopsis. Varietal names might also have been 

 proposed for the extreme variants of L. waylandvaitghani (see pi. 5, 

 fig. 5) and of L. pancanalis. For some of the species the material is 

 sufficient for a statistical analysis of the variation. Perhaps a guess 

 may be ventured that strains may be recognized. 



The variations presented by some species of Lcpidocyclina suggest 

 that the phenomena may be fundamentally similar to the variations 

 obtained by Jennings in his experimental study of DiMugia corona.^ 

 This work of Jennings should be studied by everyone who is engaged 

 in taxonomic work on foraminifera. Unfortunately, the difficulties of 

 artificially culturing the sexual, microspheric, generation of forami- 

 nifera have not yet been overcome, but of many species the asexual, 

 megalospheric, generation can be raised in numbers. E. H. Myers, 

 working at the Scripps Institution, has produced seven successive 

 generations of asexual reproduction in Discorbis globularis. 



From the accounts given in this paper of variation in single lots 

 of specimens of species of Lepidocyclhm and from work such as that 

 of Jennings, it is obvious that to attach a different specific name to 

 every variant in a lot of specimens of Lepidocyclma is an absurdity. 

 In a paper on the Tertiary larger foraminifera of Ecuador, recently 

 completed, I have pointed out that very few, if any, of the species of 



* Jennings, H. S., Heredity, variation, and the results of selection in uni- 

 parental reproduction of Difflugia corona. Genetics, vol. i, pp. 407-534, 19 figs., 

 1916. 



