Mateeials, Theie Taxonomy and Natural History 27 



L. undecimlineata Stal, L. panamensis nov. sp. from Panama, Costa Eica, and 

 the Isthmus of Darien in Colombia, and L. guatemalensis n. sp. are also quite 

 as distinct as the other two. The four are closely related and probably represent 

 species that have been differentiated from the same original stock, and which 

 are now limited to separate geographic areas. In that I do not know of inter- 

 grades between them, and that their chief differentials are sharply alternative in 

 crosses, I have retained them as species and have not classed them as geo- 

 graphical varieties. 



The relations existing between the species L. multitceniata Stal, L. oblongata 

 nov. sp., L. melanothorax Stal, L. multilineata Stal, and L. decemlineata Say 

 are confusing. As far as I can discover, L. melanothorax does not, at any 

 locality in nature, sustain itself as an independent type, but always appears 

 periodically, sometimes rather irregularly, as a recurrent sport from L. multi- 

 tceniata Stal and its geographic varieties. Upon its appearance as a sport in the 

 population, it crosses back with the parent species, within which it is at once 

 submerged. However, when isolated in a culture, it breeds true indefinitely as 

 a specific kind of living substance, and behaves in the same way that organisms 

 which are called elementary species and mutations do. The question as to what 

 L. melanothorax Stal is depends upon one's point of view and one's definition of 

 a " species." I have classed it as a " recurrent mutant." 



StaFs account of L. multitceniata and L. multilineata leaves one in doubt, 

 as far as the descriptions are concerned, as to what use should be made of the 

 two names in describing the series of forms which live on the southern half of 

 the high plateau of Mexico at the present time. Throughout the entire southern 

 end of the plateau, southward into the Rio Balsas Valley as far as Matamoras de 

 Izuca, eastward to the Plains of Apam, and northward into the upper tribu- 

 taries of the Rio Amacusac, Rio Coetzala, and Rio Atoyac, is distributed a 

 variable organism, which is unquestionably L. multitceniata Stal. Throughout 

 a considerable portion of the Oaxaca-Guerrero highlands, the Balsas Valley, 

 and up the slope of the Mexican Plateau as far as the foot of the escarpment, is 

 another species closely related to the former, but different in form, distribution, 

 in juvenile characters, as well as in life-history. In some respects this species 

 agrees with L. multilineata Stal ; but Stal considers his L. multilineata and 

 L. decemlineata Say to be one and the same species, and L. decemlineata Say is 

 not known to me in Mexico. Stal's species, multilineata, is, I believe, simply a 

 " form biotype " of L. multitceniata Stal that is fairly common in some locali- 

 ties ; at least, a biotype of L. multitceniata Stal is found which in all respects is 

 like L. multilineata Stal and therefore I have used his name for it. The form is 

 radically different from L. decemlineata Say and is distinguished by certain 

 gametic factors for body proportions from L. multitceniata Stal and its other 

 varieties, L. decemlineata Say, as recognized by Rogers and all the later Ameri- 

 can writers, is clearly distinct from the Mexican types, although in museum 

 specimens the several species might easily be confused, especially when badly 

 preserved. The living animals, however, when followed through the life-history, 

 leave not the slightest doubt as to the existence of the specific entities. 



Thorough searching of the European museums, especially the British Museum, 

 has shown interesting points that have materially helped in the solution of this 

 confused taxonomy. In the Baily collection in the British Museum there is a 



