of the Vivi'parous Aphides. 55 



parts of the einbiyo which lie in a plane passing through the 

 middle of the egg parallel to the object-slide. This opens up 

 a rich source of illusions for those who come uncritically to 

 the investigation ; but for those who are aware of the optical 

 behaviour of the pseudovitellus there arises the unconditional 

 necessity of correcting every surface-picture observed by the 

 side view corresponding to it, and, if possible, by the other 

 surface-picture (that directly opposite to the first). In the 

 former case the embryo must be turned 90°, in the latter 180°. 



Without this method of rolling, already mentioned by Harting 

 in his well-known work on the microscope, it is not possible 

 to make out the earliest development of the ^p/^w-embryo. 

 Of course it is not easy to practise the method referred to, 

 and many a fine preparation is sacrificed by clumsy handling 

 of the wire which is employed in producing the rotation. In 

 my memoir I will describe in detail the rolling method as it 

 should be constantly employed in the more delicate investi- 

 gations in insect embryology. 



I will now briefly indicate in what principal points the 

 results of my investigations differ from those obtained by the 

 distinguished Russian naturalist. 



The pseudovum possesses no chorion, but only a vitelline 

 membrane, Huxley's '^pseudo-vitelline membrane." This 

 encloses the whole contents of the %gg^ which, at a certain 

 early period (as Leuckart first remarked *), shows a distinction 

 between peripheral and central cells. We read also in 

 Huxley as follows : — " They [the pseudova] exhibit a central 

 darkish matter surrounded by a clear cortex." Upon this point 

 the sim'plest observation gives clear information. The deve- 

 lopment of the embryo now starts from the '' clear cortex," the 

 blastoderm^ which forms a multilamellar vesicle, and, indeed, 

 in this way, that at its lower pole {i. e. that turned towards 

 the vagina) a thickening is formed, from which the germinal 

 streak grows forth laterally (and near to the inner wall of the 

 blastoderm) in the form of a small thick tongue. The yelk 

 at this time contracts strongly, and places itself, as a rounded 

 mass, also at the inferior pole of the germinal vesicle. This 

 is the profile view, so to speak. If we now roll the pseud- 

 ovum through 90° we obtain a view en face ; and Metschni- 

 koff appears to liave this alone in his eye when he speaks of 

 a germinal and a vitelline " hill," the appearance of which 

 characterizes the earliest embryonic stage of the viviparous 

 Aphides. In the surface-view our glance of course falls first 

 upon the broad side of the tongue, which now looks like a 



* ' Zur Kenntniss des Generatijusweclisela und der Parthenogenese bei 

 den Insekteu ' (1868), p. 20, 



