414 THE ALIMENTARY CANAL OF THE IMAGO. 
then endeavour to state as briefly as possible the other view 
of their nature. 
(1) ‘They are developed from the proctodeum.’ This state- 
ment is usually regarded as correct. I will only observe here 
that in both the embryo and the nymph the Malpighian tubules 
communicate with the mesenteron and not with the procto- 
deum. 
Graber, although he holds the view that they are developed 
from the proctodeum, figures them in an early embryo opening 
into the mesenteron [114, Taf. III., Fig. 27], and according to - 
my observations his figure is correct. 
The statement that ‘they open into the rectum’ is incorrect. 
In nearly every insect they open into the pyloric or distal 
extremity of the chyle stomach or into the commencement of 
the proximal intestine. Inthe Blow-fly larva there are 35 mm. 
of intestine between the anus and the Malpighian ducts, and 
in the imago 6 mm. When the statement has been made, the 
authors who have made and accepted it had probably only 
those insects in mind like the Cockroach in which the distal 
intestine is very short. 
(2) ‘They appear at a very early period of development.’ 
So do the liver tubules of the vertebrate. Remak says ‘ They 
appear in the embryo chick later than the rudiment of the 
Wolffian duct, but before the first Wolffian tubules.’ 
(3) ‘ They agree anatomically and histologically with urinary 
tubules.’ 
As diverticula of the intestine high up ‘in its course, they 
certainly resemble a tubular hepato-pancreas rather than a 
kidney. The cells are large, contain fat and pigment, and 
undergo partial disintegration at least during secretion. All 
these characters point to a hepato-pancreas rather than a 
kidney. 
(4) ‘They contain crystals of uric acid.’ 
I have examined many hundreds of sections of Malpighian 
tubules, and I never yet saw any crystals in them except of 
artificial origin. In imperfect preparations numerous crystals 
are liable to form in the fat bodies, alimentary canal and blood. 
