We explore the nature of carbon-rich ( [ C/Fe ] _ { 1 D,LTE } > +0.7 ) , metal-poor ( [ Fe/H _ { 1 D,LTE } ] < –2.0 ) stars in the light of post 1D , LTE literature analyses , which provide 3D–1D and NLTE–LTE corrections for iron , and 3D–1D corrections for carbon ( from the CH G-band , the only indicator at lowest [ Fe/H ] ) .
High-excitation C I lines are used to constrain 3D , NLTE corrections of G-band analyses .
Corrections to the 1D , LTE compilations of Yoon et al .
and Yong et al .
yield 3D , LTE and 3D , NLTE Fe and C abundances .
The number of CEMP-no stars in the Yoon et al .
compilation ( plus eight others ) decreases from 130 ( 1D , LTE ) to 68 ( 3D , LTE ) and 35 ( 3D , NLTE ) .
For stars with –4.5 < [ Fe/H ] < –3.0 in the compilation of Yong et al. , the corresponding CEMP-no fractions change from 0.30 to 0.15 and 0.12 , respectively .

We present a toy model of the coalescence of pre-stellar clouds of the two populations that followed chemical enrichment by the first zero-heavy-element stars : the C-rich , hyper-metal-poor and the C-normal , very-metal-poor populations .
The model provides a reasonable first-order explanation of the distribution of the 1D , LTE abundances of CEMP-no stars in the A ( C ) and [ C/Fe ] vs. [ Fe/H ] planes , in the range –4.0 < [ Fe/H ] < –2.0 .

The Yoon et al .
CEMP Group I contains a subset of 19 CEMP-no stars ( 14 % of the group ) , 4/9 of which are binary , and which have large [ Sr/Ba ] _ { 1 D,LTE } values .
The data support the conjectures of Hansen et al .
( 63 , 59 ) and Arentsen et al .
( 11 ) that these stars may have experienced enrichment from AGB stars and/or “ spinstars ” .