THE EVOLUTION OF THE MAMMALIA. 31 
can hesitate to accept them as representing actually, or very nearly, 
the successive steps of evolutionary change and in the order in 
which those changes took place. Even with abundant material 
these series must be reconstructed only step by step, tracing each 
form back to its immediate progenitor, or forward to its imme- 
diate descendant. An unbroken series of documents is not more 
essential to the genealogist than continuous series of well-pre- 
served fossils to the palaeontologist. Neglect of this principle 
has led to no little hasty and ill considered work, which has been 
a hindrance rather than a help to progress. 
The notion that the palaeontologist must laboriously recon- 
struct his genealogical series, and must spare no pains to obtain 
perfect specimens of fossils will doubtless surprise many readers. 
It is firmly rooted in the popular mind that the palaeontologist 
needs only a single bone, or tooth, or scale, from which to recon- 
struct all the missing parts, and no pseudo-scientific illustration 
recurs more persistently in literature, in spite of repeated expos- 
ures of its utter absurdity. As a matter of fact, such ideal recon- 
structions are not of the smallest scientific value. If such easy- 
going methods could be used, palaeontology would not be worth 
attention, for it could teach us nothing. 
It is a fortunate circumstance that the well-defined genealogi- 
cal series which have already been worked out recur in many and 
widely separated groups of animals, such as mammals, reptiles, 
molluscs, brachiopods, etc. If, on comparing these different series, 
they are found to agree in their modes of evolutionary change, it 
may be inferred that such modes will prove to be of general 
application and validity. In these genealogical series each char- 
acteristic structure may be followed from its incipient stages to 
its full development, and thus it becomes easy to distinguish 
between those resemblances which are due to community of 
descent and those which have been independently acquired in 
different series. In fine, the fossils represent at least fragments, 
and sometimes very extensive ones, of the various literatures which 
register the changes of language. What is of capital import- 
ance is that these portions of the literature are preserved in the 
