A weak egg \({\mathcal E}\) of \(\text{PG}(2n+m-1,q)\) is a set of \(q^m+1\) \((n-1)\)-spaces of \(\text{PG}(2n+m-1,q)\) such that any three different \((n-1)\)-spaces span a \((3n-1)\)-space. If each element \(E\) of \({\mathcal E}\) is contained in an \((n+m-1)\)-dimensional subspace \(T_E\) of \(\text{PG}(2n+m-1,q)\) which is skew from any element of \({\mathcal E}\) different from \(E\), then \({\mathcal E}\) is called an egg of \(\text{PG}(2n+m-1,q)\). The space \(T_E\) is called the tangent space of \({\mathcal E}\) at \(E\). If \(n \not= m\), then the tangent spaces form an egg \({\mathcal E}^D\) in the dual space of \(\text{PG}(2n+m-1,q)\), the so-called dual egg of \({\mathcal E}\). The theory of eggs is equivalent with the theory of the so-called translation generalized quadrangles. The authors study eggs in \(\text{PG}(4n-1,q)\). They give a new model for eggs in which all known examples are given. They calculate the general form of the dual eggs for eggs arising from a semifield flock. Applying this to the egg arising from the Pentilla-Williams ovoid, they obtain the dual egg which is not isomorphic to any of the previous known examples. They also classify all eggs of \(\text{PG}(7,2)\) and hence determine also all translation generalized quadrangles of order \((4,16)\).