252 
On the Structure and Development of 
such a tumour in the way just described, will have no great difficulty 
in demonstrating to himself that the so-called nuclei of the spindle 
cells are, in reality, very similar both in form and size to the 
plate -like corpuscles already described in the umbilical cord. 
These facts were elicited from the study of such a sarcoma, whose 
character was definitely established by its microscopical character 
and its frequent recurrence (five times). The younger portions 
were known to be such by the description given of them by the 
patient, for, the tumours being nodular, the age of each nodule was 
pretty nearly established. In the young portions, the same plate- 
like bodies as are found in the umbihcal cord, or bodies at least 
in general very similar, were found, but they were imbedded at 
intervals in a homogeneous material in which as yet there was 
no fibrillation. This portion was actually a young growth. Older 
portions were then examined, i. e. those known clinically to be 
older. It was found that the intercellular substance had a fibril- 
lated appearance, and by suitable reagents it could be broken up 
into thin ribbons or bands, as by Miiller's fluid or by a 10 per 
cent, solution of common salt; numerous long spindle-shaped 
figures, having a flattened body at their centre, were then found ; 
and there were also similar spindle cells without any central body 
or " nucleus." Where, however, such appearances are observed it 
is easy to introduce a current, roll these bodies over, and then it 
may be seen that they are long, flattened, and of irregular size, 
appearing on profile view to Hbe spindle-shaped, and yet we may 
often press off the " nuclei " by pressing the cover upon the slide, 
showing conclusively that such spindle cells are really the inter- 
cellular substance at an early stage of fibrillation, and almost 
precisely what may be seen in certain parts of the umbilical cord 
of young embryos as already described. It does not, however, follow 
from this that all of the spindle-celled sarcomata are of these 
varieties, indeed we sometimes see them where they appear to be 
composed of real spindle-shaped bodies, closely packed together, 
and where each body contains within it a smaller flattened body. 
From a study of these gradual changes it seems likely that, in 
growth and repair, the corpuscles at first round soon become 
flattened, and have a broad envelope (b). About this envelope 
there is a further delicate and lightly attached investment, which, 
uniting with the investments of other similar bodies, is the com- 
mencement of the intercellular substance. At first the plate-like 
bodies lie in niches, as it were, in the intercellular substances, and 
if one is brushed out it leaves a socket behind it (c). As the inter- 
cellular substance increases, the corpuscles are arranged in rows and 
they become smaller, while immediately under them thin laminae 
are formed from the effused fibrin — the commencement of fibrilla- 
tion — while the corpuscles are unchanged except that they become 
