verification (a practice facilitated by Johnson's plan of merely
naming the author, without specifying the particular work quoted, or
giving any reference whereby the passage could be turned up) is
undoubtedly the reason why many of the quotations are not verbally
exact. Even so, however, they are generally adequate for the purpose
for which they are adduced, that is, they usually contain the word for
which they are quoted, and the context is more or less accurately
rendered. But in some cases it is otherwise: Johnson's memory played
him false, and he quotes a passage for a word that it does not
actually contain. As an example, under _Distilment_ he correctly
quotes from _Hamlet_, 'And in the porches of mine ears did pour the
leperous distilment.' But when he reached _Instilment_, his memory
became vague, and forgetting that he had already quoted the passage
under _Distilment_, he quoted it again as 'the leperous instilment'--a
reading which does not exist in any text of Shakspere, and was a mere
temporary hallucination of memory. There are some other curious
mistakes, which must, I suppose, have crept in either in the course of
transcription or of printing. As specimens I mention two, because they
have unfortunately perverted ordinary usage. The two words _Coco_ and
_Cocoa_--the former a Portuguese word[12], naming the _coco-nut_, the
fruit of a palm-tree; the latter a latinized form of _Cacao_, the
Aztec name of a Central American shrub, whence we have cocoa and
chocolate--were always distinguished down to Johnson's time, and were
in fact distinguished by Johnson himself in his own writings. His
account of these in the Dictionary is quoted from Miller's _Gardener's
Dictionary_ and Hill's _Materia Medica_, in which the former is spelt
_coco_ and the latter _cacao_ and _cocoa_. But in Johnson's Dictionary
the two words are by some accident run together under the heading
_cocoa_, with the disastrous result that modern vulgar usage mixes the
two up, spells the _coco-nut_, 'cocoa-' as if it were _co-co-a_, and