Key to Australian Cardiocondyla Species

The following key is based on original research as well as Seifert (2003, 2008 ).

1

 * In dorsal view the protonum with distinct humeral angles or shoulders anterolaterally => Cardiocondyla thoracica
 * In dorsal view the pronotum uniformly rounded anterolaterally => 2

2

 * Metanotal groove weakly developed or absent; postpetiolar sternite in anterodorsolateral view without prominent anterolateral corners; propodeal spines short and angular => Cardiocondyla nuda and Cardiocondyla atalanta (See Note below)
 * Metanotal groove distinct in lateral view; postpetiolar sternite in anterodorsolateral view with prominent anterolateral corners and its anterior margin appearing concave in this view; propodeal spines long and thin => 3

3

 * When viewed in lateral profile, posterior mesonotum sloping gradually into metanotal groove through a uniform curve; petiole node in dorsal view generally distinctly longer than broad; head relatively narrow (CI 72-79); sub-postpetiolar process rounded anteriorly => Cardiocondyla emeryi
 * When viewed in lateral profile, posterior mesonotum changing slope abruptly and curving steeply into the metanotal groove; petiole node in dorsal view generally slightly broader than long; head relatively broad (CI 79-86); sub-postpetiolar process angular anteriorly => Cardiocondyla wroughtonii

Note
C. atalanta was synonymised with C. nuda by Taylor (1991 ). Seifert (2003 ) removed C. atalanta from synonymy and considered it to be represented by types only, the remaining Australian material belonging to C. nuda. More recently, Seifert (2008 ) modified his concept of C. atalanta to include essentially all Australian material and found only a few specimens which belonged to C. nuda. Unfortunately he could not find any readily observable characters to separate these two taxa with recognition being based on statistical differences in the length of pilosity hairs on the first gastral tergite. He states that "gastral pubescence length, scape length and postocular distance are significantly smaller in C. atalanta while gastral pubescence density is larger than in C. nuda. However, character overlap is considerable." Thus separation of these cryptic species is not possible using a simple morphology-based key such as this and identification will require more advanced morphometric procedures such as detailed by Seifert (2008 ).