568 
the direction of the waves with the velocity of light. 
These charges lie within a cylinder meeting the plane 
of yz on the boundary of A. The magnetic vector is 
perpendicular to the electric vector and equal to it 
in magnitude both inside and outside the electric 
charges. The Maxwell-Lorentz equations are satisfied 
everywhere, and the fact that we are able to specify 
a type of radiation the total energy of which is propor- 
tional to the maximum frequency pc=» indicates that 
quantum theory may be quite compatible with these 
equations. That all frequencies up to pe occur is 
sin px _ i 
Gite 
0 
We are justified in regarding this type of field as 
elementary because, as Levi-Civita pointed out 
many years ago (Comptes vendus, t. 145, 1907), the 
electromagnetic force on the moving electricity 
vanishes everywhere and so the electricity moves 
freely under no forces and no forces are needed to 
keep it intact. In this connection it may be men- 
tioned that the force of type 2¥y(o \/1—v?/c*), which 
has been used to balance the electric force in a 
suggested model of a stationary electron (Physical 
Review, September 1922), is also zero in the present 
case because y=o and v?=c?. 
Superposing two quantum fields with 
seen at once from the equation cos qx . dq. 
_ sin (p+ dp) (*— ct) 
f and f=— sin p(#— ct) 
4—ct 
ively, 
ar , respectively 
and with coincident cylinders (or light-darts, to 
use Silberstein’s term), we obtain a wave of nearly 
homogeneous radiation of total energy tWdp. When 
the light-darts in the two fields are separate entities 
but close together, the total field still represents an 
approximately homogeneous type of radiation, but 
it also possesses some of the properties of a quantum 
radiation because the light-darts can be regarded as 
independent and one of them can be captured by an 
atom and its energy hy absorbed while the other one 
escapes. 
The composite field will behave like radiation of 
frequencies lying between v and »+dyv when reflected 
and refracted if the elementary quantum-radiation 
behaves like light of frequency »=cp. é 
To test this point we have considered the reflection 
of our elementary quantum radiation at the surface 
of the moving mirror *=ut. The reflected wave 
proves to be one of the same type as the first, the 
electric vector being —f(x-+ct)F, where f(x) = ue Ps 
We, ae 
and p’= P; ee 
Thus Doppler’s principle holds just as if pc were 
the frequency of homogeneous light instead of the 
maximum frequency contained in the quantum radia- 
tion. The energy relation e’=/v’ still holds and it 
looks as if W could be regarded as a universal constant 
he/r. 
It should be remarked that the elementary radiant 
field considered here is simply a particular case of 
a more general type of simple radiant field in which 
the rays and light-darts issue from a moving point. 
An ordinary type of electromagnetic field can be built 
up by superposing two or more simple radiant fields 
of this type and proceeding to the limit. If, then, 
ordinary electromagnetic fields are to be regarded as 
composite, there is nothing strange in regarding an 
approximately homogeneous type of radiation as 
composite. H. BATEMAN. 
California Institute of Technology, 
Pasadena, California, 
March 5. 
NO. 2791, VOL. 111] 
NATURE 

[ApriL 28, 1923 
Spermatogenesis of the Lepidoptera. 
I sHOULD be glad of the hospitality of the columns 
of NaturRE to reply to two observers whose papers 
in the December number of the Quarterly Journal of 
Microscopical Science have only recently come under 
my attention. Dr. Robert He» Bowen, of Columbia 
University, has investigated the spermatogenesis of 
the Lepidoptera, a subject which formed the first part 
of my series of papers on the ‘‘ Cytoplasmic Inclusions 
of the Germ Cells.’’ His account differs from previous 
ones chiefly in two respects—he states that the mito- 
chondrial part of the spermatid is not a skein or 
spireme but a plate-work, and what is a much more in- 
teresting objection, he denies the previous descriptions 
of the metamorphosis of this skein into a tail-sheath, 
and instead describes it as degenerating, and the tail 
region being formed of a new central substance. 
In the same Journal, Mr. Graham Cannon has re- 
described the louse mitosome, and supports Dr. 
Bowen’s conclusion that this body is not a skein but 
a plate-work. Dr. Bowen agrees with me so far as 
the general appearance of the material is concerned, 
but in a long discussion brings up a number of reasons 
for supposing that the body in question is a plate- 
work something roughly like the head of a fancy 
chrysanthemum. Mr. Graham Cannon has also given 
a similar and short account of his reasons for suppos- 
ing that this body is actually a plate-work formed by 
a system of vacuoles. ¢ 
Some years ago when Prof. Doncaster was writing 
his latest book, he came to see the material illustrating 
my view that the acrosome is always formed in 
association with the Golgi apparatus. He was shown 
my preparations of Smerinthus testes, and objected 
then to my description of the mitosome or neben- 
kern as a spireme. Dr. Bowen and Mr. Cannon will 
be glad to know this. However, I never found any 
reason to alter my views, even with such distinguished 
opposition, because it seemed to me that whether the 
““spireme ’’ was formed of a flat ribbon, or a round 
string, it was actually pulled out as the spermatid 
lengthened, much like a ball of string. The figure 
formed by the mitochondria of the spermatid is not 
a matter of importance so far as concerns the larger 
questions surrounding the study of the cytoplasmic 
inclusions. ; 
When, however, we turn to the second objection 
brought forward by Dr. Bowen, we find a matter of 
considerable importance. Dr. Bowen’s “ central sub- 
stance ’’ was believed to be the partly unravelled or 
pulled-out mitochondrial skein ; it is figured by me 
in Plate 25, fig. 47, of my paper. His account of 
this new substance being something apart from the 
nebenkern or mitosome, and of the latter not taking 
direct part in the formation of the tail, is worthy of 
reinvestigation. 
Except for Dr. Bowen’s new interpretation of 
the formation of the sperm tail, he adds nothing 
new to our knowledge of the spermatogenesis of the 
Lepidoptera. His account is valuable, however, 
because of the fact that it confirms my drawings 
of the appearances of the material already described 
by me. 
drawn from bundles which are in the process of 
formation of atypic sperms. The whole question will 
be dealt with by me in a full account elsewhere. I 
merely take this early opportunity of stating my 
position. : 
Mr. Graham Cannon’s statements will also be 
examined at length elsewhere. ; 
J. BrontiE GATENBY. 
Zoology Department, 
Dublin University, 
April 9. 
Some of his spermatid cells are effete and 
Se Pete tie 
