SCHELLHAS] WRITTEN REMAINS 597 



on Central American soil, and that many a civilized race, of which 

 not the slightest memorj' remains, existed upon that soil long before 

 the conquest. Where there is no difficulty in determining the local 

 origin of remains, as in the case of buildings and monuments, the 

 obstacles in the way of an ethnologic and chronologic determination 

 are often all the greater. 



Inductive inquiry into this ancient civilization must begin with an 

 external comparison of the remains. In this way alone can we 

 attempt to determine in how far they are of the same origin. We 

 can jjave the way to an accurate determination of the period and 

 source of separate antiquities only by means of careful sifting and 

 discrimination based on their external characteristics. 



WRITTEN REMAINS 



The written remains, to begin with these, show great uniformity. 

 We may assert positively that all the written material from Central 

 America proceeds from one and the same source: the characters are 

 essentially the same in the inscriptions, in the manuscripts, and on 

 the clay vessels and other lesser antiquities. There was but one 

 mode of writing in Central America, which emanated from one 

 center of civilization. The four manuscripts in particular are plainly 

 of one and the same origin. They may readily be divided into two 

 groups. The Troano and Cortesian codices are entirely similar, and 

 are simpler and ruder. They are undoubtedly fragments of a single 

 manuscript. The Dresden manuscript and Codex Peresianus, which 

 also strongly resemble each other, are more elegant and artistic in 

 text and pictorial representations. It is highly probable that all 

 the manuscripts pertain to one and the same nation, but Avhether 

 they belong to the same period ° is very doubtful. The forms of the 

 characters differ too nnich for us to ascribe the differences merely to 

 the peculiarities of two writers. The presumption that Codex 

 Troano-Cortesianus is the oldest lies near at hand, but it is contra- 

 dicted by the fact that not only the representiitions but also the 

 written characters in this manuscript are simi)ler. more conventional- 

 ized in form, than in the Dresden and Peresianus codices. (Jlyphic 

 characters never become more complex with time; they rather be- 

 come simplified; they liecome conventional figures, such as occur 

 repeatedly in Codex Troano-Cortesiamis (compare forms a and c. 

 figure 114, from the Dresden codex, and h and <:/, figure 114, from the 

 Troano codex). 



" Professor FOrstemann has devoted himself particularly to the question of the period 

 of the Maya inamiscrii)ts (see his Commentare zur r>resdener Haiulsiclirirt. Dresden, 

 1901; Zur Madrider Handschrift, Danzig, 1902; and Zur I'ariser Ilandsclirift, Danzig. 

 3 903). 



