46 INSTITUTE OF SOCIAL ANTHROPOLOGY—PUBLICATION NO. 4 
Sapper mentions the use of platanillo (Heliconia 
sp.) as temporary rain shelters, such as those erected 
along the trails as overnight lodgings for Indians, but 
makes no reference to any permanent thatching with 
these leaves (Sapper, 1905, p. 24). 
Ponce’s companions write in 1586 that the bamboo- 
walled house they describe, just west of Santiago 
Zambo, had a roof of “bijao”: “... the roofs of the 
houses are of leaves like those of banana, which in 
that language are called ‘bilbao,’ and with the canes 
above-mentioned, unsplit, the houses are walled.” As 
the term “bijao” is applied to various of the Musa- 
ceae, Cannaceae, and Marantaceae, positive identifi- 
cation from the common name is impossible. In all 
probability, it was a species of Calathea. 
Oviedo writes of “bihao” that the leaves “are much 
used by the Indians, especially on the mainland... . 
With these leaves they cover certain houses, and it is 
a good roofing and cleaner than grass and more beau- 
tiful from the inside of the house” (Oviedo, 1851-55, 
vol. 1, p. 276). 
Minor materials—Cane leaves of various sorts 
(sugarcane and possibly wild cane) are used in some 
localities, as at Pueblo Nuevo and San Sebastian, but 
they were invariably said to be inferior. Corn leaves 
for thatch were mentioned but once, and the report 
was vague and unconfirmed. Both the above are in- 
cluded also in Wauchope’s study (1938, p. 110), 
which cites Stephens (1841) for corn-leaf thatch. 
Maguey leaves are sometimes seen as partial roof- 
ing in the Highlands where the agave is abundant. 
They are used to cover holes in thatch, or cracks 
between boards or other covering, or to roof tem- 
porary shelters. Wauchope (1938) cites two early 
references to maguey thatch (Larrainzar, 1878, vol. 
5, p. 72, and Clavigero, 1780, vol. 2, p. 232). Maguey 
and yucca leaves are much used for thatch in 
Mexico, especially in the central valley. 
Banana leaves, though often erroneously said to 
be used in the Lowlands for thatch (undoubtedly con- 
fused with Calathea), are unsuited to roofing, for they 
rapidly shred. The only definite reference to this 
leaf for thatch is, to my knowledge, that of Blom and 
La Farge (1926-27, p. 335). 
A Ladino at Santo Domingo Suchitepequez said 
that three houses in San Gabriel had a “new kind of 
roof”? made of the split stalk of the banana plant, but 
I did not verify this. 
Durability of thatch materials.—Grass seems to 
be the preferred thatch material, and is everywhere 
said to last longer than other types. Commonly, 
Highland natives stated that is was good for 20 or 
25 years, with annual patching-up at the start of the 
rainy season. Sixteenth century estimates were more 
conservative. In the Lowlands the durability of grass 
is undoubtedly less, though at San Pedro Cutzan, 25 
years was given for grass, and 10 years for cane 
leaves (not used there). Palma del mar at Tahuesco 
is reputedly good for 20 to 25 years, though the 
estimate seems high, despite relatively low rainfall 
(about 140 cm., or 55 m., annual). Hoja de sal at 
Santo Domingo is said to last only 5 years. Standley 
states that it serves only occasionally, as “temporary 
thatching,” in the Lancetilla Valley of Honduras. 
This is probably because of the higher rainfall and 
more prolonged wet season in that area, accelerating 
disintegration more rapidly than in the Guatemala 
Pacific Lowlands, where it is an important “‘perma- 
nent” thatch. 
CRESTS 
Varied in the extreme are devices for sealing the 
ridge or apex of a thatched roof. In the case of 
pyramidal roofs, the peak is almost invariably capped 
by a large, inverted bowl called “cucuruch” (usually 
Totonicapan or San Cristébal Totonicapan pottery, 
coarse glazed ware) ; no special pottery is made for 
the purpose, regular cooking utensils being used (pl. 
2,g). It isa simple and logical solution to the prob- 
lem of closing a vulnerable point, and, as Linné has 
shown, an ancient and widespread one.°$ This author 
reproduces elaborate types used at the time of the 
Conquest, including Oviedo’s sketch, redrawn by 
Lothrop (Linné, 1938, p. 27), and, refuting Norden- 
skidld, he suggests an American origin for this trait. 
In Southwest Guatemala (1936) the distribution of 
apex pots is almost coincident with that of the square 
house. 
In those villages where square, pot-capped dwell- 
ings are built, the rectangular houses usually are 
sealed along the ridge pole by a continuous, overlap- 
ping row of large potsherds, commonly seen both in 
Santiago Atitlan and San Pedro (pl. 26, b). There 
are more of these in Santiago, which has a higher per- 
centage of square houses, probably the highest of all 
Guatemalan settlements. There are also some of this 
type in San Lucas Toliman. That the use of pot- 
sherds is not restricted to villages having both square 
and rectangular houses is evident from the rectangu- 
lar dwellings at Santa Cruz la Laguna. There, though 
68 Of the settlements visited by me, only Santa Catarina Palopé 
houses (1936) had pyramidal roofs and no ceramic caps. 
