458 
it when attacked, depends very much upon their state 
of health.”? Dr. Voelcker, the chemist of the Royal 
society, has said, ‘‘I believe the soil has a great deal 
to do with mildew. An excess of available nitroge- 
nous food appears to me to have a decided tendency 
to cause mildew in wheat. A clover-crop leaves a 
large amount of nitrogenous matter in a soil, and 
renders wheat following it liable to attacks cf rust.”’ 
Dr. Voelcker further agrees with Dr. Lawes when 
he says, in answer to Mr. Little’s letter, ‘‘ A sudden 
check by cold or continued wet weather has a decided 
tendency to favor the attacks of mildew in wheat; 
and this tendency is greater in highly manured 
Jand than in poor soil, or, at all events, on land which 
is manured with too much nitrogenous food, or on 
land naturally rich in such food.’’ Four widely dif- 
ferent soils upon which wheat had been grown were 
analyzed by Dr. Voelcker, and it was found that the 
amount of mildew determined by extended observa- 
tions varied directly with the per cent of nitrogenous 
matter in the soil. But much depends on previous 
cropping, and therefore the ratio between mildew 
and nitrogenous matter in the soil may vary to a 
limited extent. 
The large amount of evidence gathered, and pre- 
sented in extended tables, shows that some sorts are 
more capable than others of resisting rust, though no 
varieties are rust-proof. White wheats suffer more 
than red sorts. It is best to sow early maturing 
varieties, and sow them early. 
Byron D. HALSTED. 
New York. 
THE CODEX CORTESIANUS. 
Codex Cortesianus, manuscrit hicratique des anciens 
indiens de l’ Amcrique Centrale conservé du Musée 
archaeologique de Madrid photographié et publié 
pour la premiere fois avec une introduction et une 
vocabulaire de l’écriture hicratique Yucateque. Par 
Lton pvE Rosny. Paris, Maisonneuve, 1883. 
26+ 49 p.,42 pl. 4°. 
Tis volume by Léon de Rosny is undoubt- 
edly the most important contribution to Cen- 
tral-American paleography which has appeared 
since the publication of Landa’s ‘ Relacion,’ 
and the ‘ Manuscrit Troano’ by Brasseur: de 
Bourbourg. In it we have a photo-engraved 
reproduction of the recently found aboriginal 
manuscript known as the ‘ Codex Cortesianus,’ 
thus adding one more to the brief list of pre- 
Columbian Maya documents which have so far 
been discovered. The name ‘ Cortesian’ has 
been applied to it because of the supposition 
that it had once belonged to Hernando Cortez. 
Up to 1876 but three of these manuscripts — 
the ‘ Dresden codex,’ the ‘ Codex Troano,’ and 
the ‘ Codex Peresianus’ (or ‘ Manuscrit mexi- 
cain No. 2’)—had been brought before the 
public. About this time a proposition was 
made to the Bibliothéque impériale of Paris 
by some one in Spain (the name is not given) 
SCIENCE. 
_ obtain a copy of it. 
[Vou. IIL, Now 620 aN 
to sell to it an ancient American manuscript. 
A photographic copy of two pages accom- 
panied the proposition as specimens of the vol- 
ume. On account of the high price demanded, 
the proposition was not accepted. Shortly 
afterwards it was obtained by the Spanish goy- 
ernment, and deposited in the archeological 
museum at Madrid. One of these two pages 
was copied by Mr. Rosny in plate 11 of his 
‘¢ Hssai sur le déchiffrement de I’ écriture hié- 
ratique de l’ Amérique Centrale;’’ and the 
other, which is beyond question a missing half . 
of the initial page of the ‘ Codex Troano,’ in 
plate 5 of his ‘* Documents écrits de V antiquiteé 
américaine.”’ 
In 1880 Mr. Rosny went to Madrid expressly 
to see and study this codex, and, if possible, to 
Through the kindness of 
Don Juan de Dios de la Rada, the curator of the 
museum ; his mission was eminently successful, 
as he was permitted, not only to examine it, but 
to make two complete photographic copies of it. 
It was from these, I presume, that the plates 
of the present work were made. 
We learn from the introduction, that the 
original, like the other three Maya manu- 
scripts, is written on both sides of a strip 
(probably of Maguey paper) covered with a 
coat of white paint. Judging by the specimen 
given in Mr. Rosny’s ‘ Essai sur le déchiffre- 
ment,’ plate 11, I presume the figures are par- 
tially colored, though not so highly nor to the 
same extent as in the Troano manuscript ; but. 
unfortunately this is not shown in the present: 
work. 
The general appearance, the figures, the form 
of the characters, and numerous other particu- 
lars, prove very clearly that it is more closely 
related to the Troano manuscript than to any 
other one of the Central-American codices. 
This is so apparent, that Mr. Rosny has sug-. 
gested that the two are parts of one original. 
work. The fact that we find here the missing 
half (by this we know that one-half is missing) 
of the ‘ titlepage’ of the Troano manuscript 
is a very strong argument in favor of this view. 
Still, I am disposed to doubt its correctness, for 
the following reasons: 1°. On plates 39 and 40, 
upper division, we find an exact repetition of 
the five figures in the top division of plates 
29 and 30 of the Troano manuscript; 2°. In 
the plates of the latter half, quite a number 
of numerals are introduced into the text, and 
joined to characters to which they are never 
attached in the manuscript; 3°. The form of 
the serpent-figures (no one can fail to remark _ 
the strong resemblance between the heads of: 
some of these serpent-figures and the dragon- 
