1891.] Botany. 581 
(preferably in Latin) according to phytographic rules.. Indeed, it is 
very difficult and often very uncertain to seek out in the multitude of 
details the essential and distinctive characters. 
2. The diagnosis is with certain authors. (especially among the 
cryptogamists) extraordinarily detailed and prolix; in others too 
laconic. A good diagnosis should give in a clear and careful state- 
ment only the essential and distinctive characters; all observations 
concerning details should be given after the diagnosis; for new species 
it is necessary to give the relationships with the nearest related known 
Species. Whoever determines new species knows how much time it 
takes if he has to do with very prolix diagnoses without reference to 
relationships. 
3- Experience has already shown, at least for the Cryptogams, that 
relative to authority it is very useful to give in brackets the author who 
first described the concerned species under other genera. It is always 
necessary to subjoin the name of the author who has transferred the 
Species from the original genus to another, for otherwise one must 
assume that the writer of the treatise in which the combination of names 
is cited is also the author of this combination. We find, for example, 
in Winter’s works, names as Keine: “‘Spherella convexula (Schwein.), 
Syn. Spheria convexula Schwein.” If the name of Thiimen is not 
added after the parentheses we must’assume that Winter i is the author of 
the combination, and then we should have, according to the rules of 
other botanists, the two expressions: Spherella convexula (Schwein. ) 
Wint., or Spherella convexula Wint., both of which are incorrect. 
But if we say Spherella convexula (Schwein.)Thiim., we have the 
entirely new information that Schweinitz established the species and 
Thiimen transferred it to the proper genus. 
4. In the description of parasitic Cryptogams the host plants (or 
animals) are to be given with their technical Latinized nomenclature, 
The common names (English, Italian, German, etc.) are often difficult 
to identify. 
- In measuring organs, microscopic or macroscopic, one system, 
the metric, should be employed ; for microscopical measurements the 
Micromillimetre, or mikron (#) is suggested in place of fractions. 
The different measures and fractions are often the source of error or 
d 
doubt. 
6. For concise statements of the dimensions of microscopic organs 
it is suggested (as it is moreover done in manifold ways) to write first 
the figure of the length, then that of the greater breadth, the two 
connected by the sign ~, and the character » omitted ; for flat organs 
