NO.    5      MALE  GENITALIA  OF  ORTHOPTEROID  INSECTS SNODGRASS         1 1 
If  the  acridid  phallus  is  a  direct  product  of  the  united  appendages 
of  the  tenth  abdominal  segment,  a  similar  origin  for  the  organ  has 
not  been  observed  in  any  other  insect.  The  embryonic  appendages 
of  the  tenth  segment  are  known  otherwise  to  be  retained  only  as 
larval  "  legs  "  in  Neuroptera,  Trichoptera,  Lepidoptera,  and  lower 
Hymenoptera.  The  phallic  rudiment  of  Lepidoptera  is  said  by  Mehta 
(1934)  to  appear  during  the  fourth  larval  instar  as  a  small  conical 
outgrowth  in  the  base  of  an  ectodermal  genital  pouch  formed  earlier 
on  the  venter  of  the  ninth  abdominal  segment.  Toward  the  end  of 
the  larval  period  the  single  rudiment  splits  into  a  pair  of  phallic  lobes, 
which,  before  pupation,  come  together  and  unite  about  a  central  de- 
pression of  the  integument  that  becomes  the  unpaired  part  of  the 
ejaculatory  duct.  The  final  development  of  the  phallus  takes  place 
in  the  pupal  stage.  In  Mehta's  account  there  is  certainly  nothing  to 
suggest  any  possible  relation  of  the  primitive  phallic  lobes  to  the  larval 
appendages  of  the  tenth  segment.  In  Blattidae,  there  may  be  three 
distinct  phallic  lobes  (fig.  15  B),  and  in  Tettigoniidae  as  many  as 
six  in  a  young  nymph  (fig.  2;^  B).  We  may  question,  therefore, 
whether  the  two  primary  phallic  lobes  of  Acrididae  are  not  inde- 
pendent outgrowths  of  the  genital  integument  that  might  be  con- 
fused with  the  simultaneously  disappearing  appendage  rudiments  of 
the  tenth  abdominal  segment. 
The  phallic  components  have  usually  been  attributed  on  theoretical 
grounds  to  the  appendages  of  the  ninth  abdominal  segment,  sup- 
posedly endites  or  "  endopodites  "of  these  appendages,  thus  making 
the  male  organ  partly  equivalent  to  the  female  ovipositor ;  but  of  this 
there  is  certainly  no  evidence  from  ontogeny.  It  is  shown  by  Mehta 
(1934)  that  the  phallic  lobes  of  Lepidoptera  have  no  anatomical  re- 
lation to  the  integumental  outgrowths  of  the  ninth  segment  that  give 
rise  to  the  clasping  appendages,  or  valvae,  of  the  adult  genital  ap- 
paratus. It  seems  most  probable,  therefore,  that  the  insect  phallus 
is  either  an  independent  product  of  the  genital  integument  around  the 
mouth  of  the  ejaculatory  duct,  or  a  product  of  the  tenth  pair  of 
abdominal  appendages.  If  the  primary  phallic  components  really 
are  lobes  of  the  tenth  abdominal  segment  that  converge  from  a  lateral 
position  and  join  with  each  other,  it  might  be  supposed,  judging  from 
the  facts  in  other  arthropods,  that  these  lobes  are  primitive  paired 
penes  derived  from  the  limb  bases,  rather  than  that  they  are  the 
appendages  themselves,  for  in  no  case  does  a  genital  duct  traverse  an 
entire  appendage.  The  original  presence  within  the  lobes  of  the 
terminal  ampullae  of  the  primary  ducts  would  suggest  such  an  in- 
terpretation, but  in  this  case  it  is  evident  that  the  ampullae  leave  the 
